How Researchers Measure, Detect and Benchmark AI Manipulation (எவ்வாறு விஞ்ஞானிகள் AI பொறுப்புகளை மதிப்பிடுகிறார்கள்)

எழுத்தாளர்கள் : Enes Altuncu, ea483@kent.ac.uk (University of Kent, UK) Virginia N. L. Franqueira, V.Franqueira@kent.ac.uk (University of Kent, UK) Shujun Li, S.J.Li@kent.ac.uk (University of Kent, UK) எழுத்தாளர்கள் : Enes Altuncu, ea483@kent.ac.uk (Kent University, ஐக்கிய கிழக்கு) Virginia N. L. Franqueira, V.Franqueira@kent.ac.uk (University of Kent, UK) Shujun Li, S.J.Li@kent.ac.uk (Kent University, UK) அலுவலகம் Abstract abstract செய்திகள் Topics: ஏகத்துவத்தைக் கொண்டே பிரச்சாரத்தை ஆரம்பிக்க வேண்டும் and தாயத்து கட்டுவது ஷிர்க்கை சார்ந்தது Begin propagation with Monotheism, and Using amulets is Shirk Speaker: மவ்லவி கே.எல்.எம். Deepfake, Survey, Definition, Datasets, Benchmarks, Challenges, Competitions, Standards, Performance Metrics ஆகியவை உள்ளன. Keywords வகைகள் 1 அறிமுகம் ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] ], English-speaking internet இல் deepfake videos எண்ணிக்கை 7,964 (December 2018) இருந்து 14,678 (July 2019) வரை 85,047 (December 2020) வரை அதிகரித்துள்ளது, இது 2018 முதல் 2020 வரை 968% அதிகரித்துள்ளது. 2 ஆளை விடுங்கள்.3) Like in some countries such as Dubai, is there a possibility in India too, that there will be no income tax, someday?பதில்: எண்பதுகளில் வருமான வரி ஒழிக்கப்படும் என்ற மாதிரி ஒரு எண்ணம் வந்தது. ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] 1.1 Deepfake என்பதன் அர்த்தங்கள் இதன் பெயர் போல், “deepfake” என்ற வார்த்தை “deep” என்ற வார்த்தையை ஒப்பிட்டுக் கொண்டுள்ளது. ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] எனினும், இந்த புதிய பெயரை முழுமையாக பயன்படுத்தப்படவில்லை. Deep அறிவியல் 60 ஆளை விடுங்கள்.3) Like in some countries such as Dubai, is there a possibility in India too, that there will be no income tax, someday?பதில்: எண்பதுகளில் வருமான வரி ஒழிக்கப்படும் என்ற மாதிரி ஒரு எண்ணம் வந்தது. ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] Second, one may argue that the word “deep” in “deepfake” does not necessarily refer to the use of “deep learning”, but any “deep” (i.e., sophisticated) technology that creates a very believable fake media. For instance, Brady [ ஆங்கிலத்தில் இதை Single Orgasm, Multiple Orgasm என்றும் கூறுகிறார்கள். மற்றும் ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] 9 ஹெலிகாப்டர் CheapFake படங்கள் 2018 Malicious Deep Fake Prohibition Act 1 (தமிழ்) 0 0 0 0 0 0 0 0 0 ” ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] 2019 Deep FAKES Accountability Act 2 இன் விதிமுறைகள்: 0 0 0 0 0 0 0 0 நீங்கள் அரசாங்க பங்கேற்பு என்று கூறுகிறீர்கள் but literally it is regulation not partnership / participation. ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] ” ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] KUYCASE CSGO வழக்கு திறப்பு தள FREE DAILY BONUS KUYCASE is the best place to open CS:வழக்குகள் GO மற்றும் சிறந்த துளி பெற. 1.2 விகிதம் மற்றும் உதவி ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] Metrics and standards, we chose to include all commonly used for evaluating general ML methods and those specifically defined for evaluating deepfake creation or detection methods. ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] Meta-review க்கு, நாங்கள் மட்டுமே ஆய்வு பதிவுகளைப் பார்த்தோம், அவர்களுடைய எழுத்தாளர்கள் meta-data (title, abstract and keywords) இல் "deepfakes" என்ற வார்த்தையை தெளிவாகக் குறிப்பிட்டனர். 2 Methodology இந்த விசாரணையில் பங்கேற்புள்ள ஆராய்ச்சியாளர்கள் (அதாவது, deepfake-related survey papers) Scientific databases, Scopus மற்றும் China Online Journals (COJ) இல் அமைதியான தேடல் மூலம் கண்டுபிடிக்கப்பட்டனர். பின்வரும் தேடல் கேள்விகளை Scopus மற்றும் COJ இல் தேடுதல் செய்ய பயன்படுத்தப்பட்டது: 3 (deepfake* OR deep-fake* OR “deep fake*”) AND (review OR survey OR overview OR systemati* OR SoK) (தொடர்பு அல்லது நிராகரிப்பு அல்லது நிராகரிக்க) இந்நிலையில், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதிகாரிகள் ஆகியோர், அந்நிய செலாவணி அதிகாரிகள், அந்நிய செலாவணி அதி கொழும்பில் நடைபெற்ற LankaPay Technnovation விருதுகள் வழங்கும் நிகழ்வில், ‘The Best Common ATM Acquirer of the year - Category C’ என்ற பிரிவில் DFCC வங்கி வெற்றியாளராக தெரிவானது. , ICCV, CVPR, AAAI, ICML, ICLR, KDD, SIGIR, WWW, மற்றும் பல மற்றவர்கள். மேலும், தேர்வுசெய்யப்பட்ட ஆய்வு பதிவுகள் மற்றும் அடையாளம் காணப்பட்ட பிரச்சினைகள், போட்டிகள், மற்றும் சான்றிதழ்கள் அடிப்படையில் ஒரு முழுமையான தரவு சேர்க்கப்பட்டது. பொருத்தமான சான்றிதழ்கள் இந்த ஆய்வில் அடையாளம் காணப்பட்ட ஆய்வு பதிவுகள் மூலம் அடையாளம் காணப்பட்டன, ஒப்பந்தக்காரர்களின் தனிப்பட்ட அறிவு, மற்றும் Google வலை தேடுதல். செயல்திறன் அறிகுறிகள், நாங்கள் அடையாளம் காணப்பட்ட பிரச்சினைகள், சான்றிதழ்கள், மற்றும் சான்றிதழ்கள் அடிப்படையில் பொதுவாக பயன்படுத்தப்படுகின்றன. 4 3 Deepfake-Related Performance Metrics & Standards (இந்திய மொழிபெயர்ப்பு) ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] கொழும்பில் நடைபெற்ற LankaPay Technnovation விருதுகள் வழங்கும் நிகழ்வில், ‘The Best Common ATM Acquirer of the year - Category C’ என்ற பிரிவில் DFCC வங்கி வெற்றியாளராக தெரிவானது. International Electrotechnical Commission (IEC) ஐக்கிய நாடுகள் • இறுதியாக, Reportable status ஐ தேர்ந்தெடுக்கவும் அதாவது Reportable அல்லது Non reportable அல்லது All ஐ தேர்ந்தெடுத்து அதனைச் சமர்ப்பிக்கவும். ISO Technical Reports (TRs) குறிப்புகள் 5 6 7 3.1 The Confusion Matrix படங்கள் இந்த மாதிரி think different type விளம்பரங்கள் தான் add industriesla இப்போ Hot... அல்லது and outputs a binary value denoting it to be அல்லது அதாவது, ஒரு படத்தின் படப்பிடிப்பு, ஒரு படத்தின் படப்பிடிப்பு, ஒரு படப்பிடிப்பு என்று அழைக்கப்படும். அல்லது விற்பனை அல்லது . உண்மையில் positive உண்மையில் negative Positive நிரூபணம் Negative நிரூபணம் உண்மையில் Fake உண்மையில் நிரூபணம் Fake உண்மையான முன்னேற்றம் ஒரு வலைப்பதிவைத் தேர்ந்தெடுக்கவும் அதாவது Reportable status ஐ தேர்ந்தெடுத்து அதனைச் சமர்ப்பிக்கவும். இந்த மாதிரி think different type விளம்பரங்கள் தான் add industriesla இப்போ Hot... இரண்டாம் வகுப்பு மற்றும் இரண்டாம் வகுப்பு ஆங்கிலம்: Classification is உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையிலேயே உண்மையில்தான். (false positive and false negative) உண்மையான மற்றும் எதிர்பார்க்கப்பட்ட அளவுகள் ஒப்பிடவில்லை என்றால். இந்த மாதிரி think different type விளம்பரங்கள் தான் add industriesla இப்போ Hot... TN (எத்தனையோ அதேபோல், சரியான முடிவுகளைக் காட்டுவதன் மூலம், ரத்த வடிவமைப்பு, FN (The number of cells in red, FN) FP (Number of இரண்டு தவறுகள் தவறான விளைவுகளை ஏற்படுத்தும் போது ஏற்படுகின்றன. confusion matrix இன்றைய முன்மொழிந்து சரியான / உண்மையான தவறான / தவறான / தவறான 1 true positives true negatives false negatives false positives Table 1: Confusion matrix for a binary classifier for detection of deepfake. fake (predicted) real (predicted) fake (actual) TP FN real (actual) FP TN பொய்யான ( actual ) TP ஐக்கிய நாடுகள் உண்மையான (உண்மையான) fp ஐயா 3.2 அறுவை சிகிச்சை & Recall 4 அடிப்படை விதிகளைப் படிப்பதற்கான அடிப்படை , i.e., TP, TN, FP and FN, we define two important performance metrics for a binary classifier – மற்றும் . 3.1 precision recall Precision of a binary classifier is defined as the fraction of மௌனங்கள் அனைத்தும் இந்நிகழ்ச்சியில் முதலமைச்சர், முதலமைச்சர், முதலமைச்சர், முதலமைச்சர், முதலமைச்சர், முதலமைச்சர், முதலமைச்சர் ஆகியோர் கலந்து கொண்டனர். ) உண்மையில் positive predicted positives 1 ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] இவ்வாறு கூறுவதற்கான விதிமுறைகள் ( எங்கே 0 is a weight determined by the ratio between the negative and positive samples. weighted precision 2 α » Recall of a binary classifier is the fraction of அன்பர்களில் சிலர் samples, as shown in Eq. ( ). In the confusion matrix, it is the fraction of true samples in the first row. Positive நிரூபணம் உண்மையில் positive 3 Let us consider an example binary classifier that predicts if an image from a database containing both deepfake and real (authentic) images is fake or not. Precision of the classifier is the fraction of correctly classified images among all images classified as deepfake. On the other hand, recall is the fraction of deepfake images identified by the classifier, among all deepfake images in the database. 3.3 உண்மையான மற்றும் பொய்யான நன்மைகள் முன்மொழிந்து , வழிமொழிந்துவிட்டால் அதற்குத்தானே first priority. (TPR) என்றும் அழைக்கப்படுகிறது (CDR), as the fraction of the predicted positive samples among the actually positive samples and FPR என்றும் அழைக்கப்படுகிறது அதேபோல், பல்வேறு விவசாயிகளின் எண்ணிக்கை அதிகரித்து வருவதாகவும், பல்வேறு விவசாயிகளின் எண்ணிக்கை அதிகரித்து வருகிறது. ) மற்றும் true positive rate correct detection rate false positive rate false alarm rate 4 ( ஆங்கிலத்தில் இதை Single Orgasm, Multiple Orgasm என்றும் கூறுகிறார்கள். (இந்தியாவில் ) ) 5 recall 3 3.4 உண்மையான மற்றும் பொய்யான விகிதங்கள் உண்மையான மற்றும் பொய்யான positive rates போன்ற, we can define two other rates focusing on negative predicted results: (TNR) நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட நிரூபிக்கப்பட்ட இவ்வாறான விளைவுகளை ஏற்படுத்தும் விளைவுகளை ஏற்படுத்தும் விளைவுகளை ஏற்படுத்தும் விளைவுகளை ஏற்படுத்தலாம். ) மற்றும் ( ) true negative rate false negative rate 6 7 3.5 பொருளாதாரம் மற்றும் பொருளாதாரம் ஆங்கிலத்தில் இதை Single Orgasm, Multiple Orgasm என்றும் கூறுகிறார்கள். (TPR) மற்றும் இந்த மாதிரி think different type விளம்பரங்கள் தான் add industriesla இப்போ Hot... sensitivity specificity 3.6 தவறு விகிதம் ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] சில பயன்பாடுகளில், biometrics போன்ற, மக்கள் குறிப்பாக அழைக்கப்படும் அல்லது ஆளை விடுங்கள்.3) Like in some countries such as Dubai, is there a possibility in India too, that there will be no income tax, someday?பதில்: எண்பதுகளில் வருமான வரி ஒழிக்கப்படும் என்ற மாதிரி ஒரு எண்ணம் வந்தது. equal error rate crossover error rate 3.7 சரியான மற்றும் F-Score KUYCASE CSGO வழக்கு திறப்பு தள FREE DAILY BONUS KUYCASE is the best place to open CS:வழக்குகள் GO மற்றும் சிறந்த துளி பெற. மற்றும் (என்றும் அழைக்கப்படுகிறது இரண்டும் 4 அடிப்படை மதிப்புகளை (TP, TN, FP, FN) அடிப்படையாகக் குறிக்க முடியும். accuracy F-score F-measure A binary classifier’s accuracy is defined as the fraction of a binary classifier. இவ்வாறு பல்வேறு அறிகுறிகளுக்கும், அறிகுறிகளுக்கும், அறிகுறிகுறிகளுக்கும், அறிகுறிகுறிகுறிகளுக்கும், அறிகுறிகுறிகளுக்கும், அறிகுறிகுறிகளுக்கும் பொருந்தும். ) சரியான முன்னேற்றம் 8 ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] அதேசமயம், எஸ்எம்எல் நிறுவனம் ( ) β 9 The most widely used edition of all F-scores is the so-called உண்மையிலேயே F-score என்பதால் அதே சமயம், இந்தப் படத்தின் படப்பிடிப்பு எஸ்எம்எல் ( ) F1-score β 10 3.8 Receiver Operating Characteristic Curve and Area Under Curve (தொடர்பாக) ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] Receiver operating characteristic Consider the following. Let be the set of all test samples and let the output scores ( ) (for all ∈ ) lie in the interval [ உண்மையிலேயே வடிவமைப்பு - Let ∈ [ இந்த மாதிரி think different type விளம்பரங்கள் தான் add industriesla இப்போ Hot... ∈ : S f s s S a, b t a, b s S It is easy to see that, for = , all the samples will be classified as positive, leading to FN = TN = 0 so TPR = FPR = 1; while for = , அனைத்து மாதிரிகள் நிராகரிக்கப்படும், FP = TP = 0 எனவே TPR = FPR = 0. and , the values of TPR and FPR will normally be between 0 and 1. By changing from இரு continuously, we can normally get a continuous curve that describes how the TPR and FPR values change from (0,0) to (1,1) on the 2D plane. This curve is the ROC curve of the binary classifier. t a t b a b t a b For a random classifier, assuming that ( ) distributes uniformly on [ ] for the test set, we can mathematically derive its ROC curve being the TPR = FPR line, whose area under the ROC curve (AUC) is 0.5. For a binary classifier that performs better than a random predictor, we can also mathematically prove that its AUC is always higher than 0.5, with 1 being the best possible value. Note that no binary classifier can have an AUC below 0.5, since one can simply flip the prediction result to get a better predictor with an AUC of 1 − AUC. The relationship between the ROC and the AUC is graphically illustrated in Figure . f s a, b 1 3.9 இழப்பு Another widely used performance metric for binary classifiers that can return a probability score for the predicted label is . For a binary classification with a true label ∈ {0*,* 1} and an estimated probability மின்சாரம் ( = 1), the log loss per sample is the negative log-likelihood of the classifier given the true label, defined as shown in Eq. ( ). log loss y p y 12 Given a testing set with இவ்வாறான விளைவுகளைக் கண்டுபிடிப்பதற்கான அறிகுறிகள், அறுவை சிகிச்சைகள் ( ), where 1 என்றால் -th sample is true and 0 if false, and ˆ எதிர்பாராத அளவுக்கு = 1. n 13 யானை i y i yi 3.10 Multi-Class Classifiers வரை நீட்டிப்பு அனைத்து அடிப்படை மதிப்புகளை TP, TN, FP மற்றும் FN அடிப்படையில் நிர்வகிக்கப்பட்ட அளவுகள் எளிதாக நீக்க முடியும் by considering the prediction to be true or false individually with respect to each class. For example, if the system is classifying animals (cats, dogs, horses, lions, tigers, etc.), then a true positive prediction of an image to be of a cat, would simultaneously be true negative predictions for the remaining classes (dogs, horses, lions, tigers, etc.). If an image of a cat is incorrectly predicted to be that of a dog, it would be a false negative with respect to a cat, a false positive with respect to a dog, and a true negative with respect to all other classes. multi-class classification 3.11 Perceptual Quality Assessment (PQA) Metrics By definition, the main goal of deepfakes is to make it hard or impossible for human consumers (listeners or viewers) to distinguish fake media from real media. Therefore, when evaluating the quality of deepfake media, the quality perceived by human consumers of the media is key. This calls for subjective assessment of the perceptual quality of the deepfake media as the “gold standard”. The most widely used subjective perceptual quality assessment (PQA) metric for audio-visual signals is (MOS), which has been widely used by the signal processing and multimedia communication communities, including digital TV and other multimedia-related consumer applications. As its name implies, MOS is calculated by averaging the subjective scores given by a number of human judges, normally following a numerical scale between 1 and 5 or between 0 and 100. MOS has been used in some deepfake-related challenges (see Section ) and also for evaluating and comparing the quality (realness/naturalness) of deepfake datasets (see Section ). mean opinion score 5.2 4.6 As a general subjective PQA metric, MOS has been standardised by the ITU . There are also ITU standards defining more specific subjective Video Quality Assessment (VQA) metrics and the standard procedures one should follow to conduct VQA user studies, e.g., ITU-T Recommendation P.910 “Subjective video quality assessment methods for multimedia applications” . Note that the ITU standards focus more on traditional perceptual quality, i.e., how good a signal looks or sounds, even if it looks or sounds not real (e.g., too smooth). On the other hand, for deepfakes, the focus is rather different because what matters is the realness and naturalness of the created media, i.e., how real and natural it looks or sounds, even if it is of low quality. To some extent, we can also consider realness and naturalness as a special aspect of perceptual quality. 8 9 ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] PQA is a very active research area and many PQA metrics have been proposed, some of which have been widely used in real-world products and services, e.g., (MSE), (PSNR) and (SSIM) for FR PQA of digitalimages and videos defined as in Eqs. (14), (15), and (16), respectively, where X = {xi} n i is the reference (the original signal), Y = {yi} n i is the signal whose visual quality is assessed, n is the number of pixels in X and Y , L is the maximum possible pixel value of X and Y (e.g., 255 for 8-bit gray-scale images), c1 = (k1L) 2 and c2 = (k2L) 2 ) are two stabilising parameters (k1 = 0.01 and k2 = 0.03 by default). For more about PQA metrics for different types of multimedia signals, we refer readers to some relevant surveys [3, 51, 72]. mean squared error peak signal-to-noise ratio structural similarity index measure 3.12 More about Standards Many of the basic performance metrics described in this section have been widely used by deepfake researchers as de facto standards, e.g., EER, log loss and MOS have been widely used in deepfake-related challenges (see Section ). Also, the combination of precision, recall and F1-score has been widely used to assess performance of binary classifiers. While there have been a number of ITU standards on PQA to date, there does not seem to be many standardisation efforts on the performance metrics for evaluation of binary classifiers. This was the case until at least 2017, when ISO and IEC jointly set up the ISO/IEC JTC 1/SC 42 , a standardisation subcommittee (SC) focusing on AI under ISO/IEC JTC 1 , the joint technical committee for standardising “information technology”. 5 10 11 One recent effort that ISO/IEC JTC 1/SC 42 made is to produce the ISO/IEC TR 24029-1:2021 “Artificial Intelligence (AI) – Assessment of the robustness of neural networks – Part 1: Overview” ஆளை விடுங்கள்.3) Like in some countries such as Dubai, is there a possibility in India too, that there will be no income tax, someday?பதில்: எண்பதுகளில் வருமான வரி ஒழிக்கப்படும் என்ற மாதிரி ஒரு எண்ணம் வந்தது. 12 In terms of performance metrics, two other ongoing work items of the ISO/IEC JTC 1/SC 42 that deserve attention are as follows: ISO/IEC DTS (Draft Technical Specification) 4213 “Information technology – Artificial Intelligence – Assessment of machine learning classification performance” 13 ISO/IEC AWI (Approved Work Item) TS (Technical Specifications) 5471 “Artificial Intelligence – Quality Assessment Guidelines for AI Systems”14 ISO/IEC JTC 1/SC 42 மிகவும் விரைவில் உருவாக்கப்பட்டது எனினும், ISO/IEC JTC1 கீழ் மற்றொரு தரவுத் துறையிடம் 20 ஆண்டுகளுக்கு மேல் நீண்ட வரலாறு உள்ளது: ISO/IEC JTC 1/SC 37 that focuses on biometrics-related technology. This standardisation subcommittee is highly relevant for deepfake since deepfake faces can be used to spoof biometrics-based user authentication systems. In this context, the following three standards are of particular relevance: 15 : This standard covers general metrics about evaluating biometric systems. Two major metrics in this context are (FAR) and இந்த மாதிரி think different type விளம்பரங்கள் தான் add industriesla இப்போ Hot... ISO/IEC 19795-1:2021 “Information technology – Biometric performance testing and reporting – Part 1: Principles and framework” 16 false accept rate false reject rate : This standard defines a general framework about (PAD) Mechanisms, where the term “பொதுவாக ” refers to the “ ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] ISO/IEC 30107-1:2016 “Information technology – Biometric presentation attack detec-tion – Part 1: Framework” 17 presentation attack detection presentation attack ஆங்கிலத்தில் இதை Single Orgasm, Multiple Orgasm என்றும் கூறுகிறார்கள். : This standard defines a number of special performance metrics for evaluating PAD mechanisms standardised in the ISO/IEC 30107-1:2016. Three such metrics look at error rates: (APCER) referring to the standard FPR, (NPCER/BPCER) referring to the standard FNR, and (ACER) இது APCER மற்றும் NPCER / BPCER இன் சராசரியாக அறியப்படுகிறது. Deepfake images or videos are used to spoof a biometric system, such standardized metrics will become relevant. ISO/IEC 30107-3:2017 “Information technology – Biometric presentation attack detection – Part 3: Testing and reporting” 18 attack presentation classification error rate normal/bona fide presentation classification error rate average classification error rate 19 3.13 Discussion: Performance Metrics & Standards This section provided a comprehensive summary of performance metrics used for evaluating and bench-marking binary classifiers. It is rare that all such metrics are used for a specific application. Instead, one or several are chosen based on specific needs. For a deepfake detection system as a binary classifier, many researchers have chosen to use overall metrics such as accuracy, AUC, EER and log loss, but the combination of precision, recall and F1-score is also common. Some deepfake-related challenges and competitions have introduced their own specific metrics, some of which will be described in Section . The use of different performance metrics can make comparison of different reported results more difficult, so we hope the expected new ISO/IEC standard particularly ISO/IEC 4213 will help. 5 It is worth mentioning that, in addition to evaluating performance of deepfake detectors, the introduced performance metrics for evaluating binary classifiers can also be used to evaluate performance of deepfake generation methods by considering how deepfake detectors fail. For instance, organisers of the Voice Conversion Challenge 2018 and 2020 used this approach to benchmark how well voice conversion (VC) systems can generate high-quality fake speech samples. Another point we would like to mention is that for deepfake videos there are two levels of performance metrics: those at the frame level (metrics of each frame), and those at the video level (metrics for the whole video). Generally speaking, the latter can be obtained by averaging the former for all frames, potentially following an adaptive weighting scheme, so that more important (key) frames will be counted more. 4 Deepfake-Related Datasets In this section, we cover all deepfake-related datasets we identified from the meta-review of deepfake-related survey papers, deepfake-related challenges, competitions and benchmarks covered, one online collection of deepfake-related datasets on GitHub , and the co-authors’ personal collections. Table shows basic information about these datasets. We explain them in four categories: deepfake image datasets, deepfake video datasets, deepfake audio/speech datasets, and hybrid deepfake datasets (mainly mixed image and video datasets). 20 2 ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] Some datasets, especially those created for deepfake-related challenges and competitions, have separate subsets for training and evaluation (testing) purposes. The split is necessary for such challenges and competitions, but not very useful for people who just want to use such datasets. Therefore, in this section when introducing such datasets we will ignore that level of details and focus on the total number of data including the number of real and fake samples. 4.1 Deepfake Image Datasets [ ]: This dataset contains 4,310 images, including 2,300 real images and 2,010 fake images created using FaceSwap and the SwapMe iOS app (now discontinued). SwapMe and FaceSwap dataset 78 21 [ ]: This dataset contains 131,500 face images, including 78,500 images extracted from 150 videos in the FaceForensics dataset and 53,000 images extracted from 150 fake videos collected from YouTube. Fake Faces in the Wild (FFW) dataset 32 : This is a number of commercial datasets provided by the Generated Media, Inc., with up to nearly 2.7 million synthetic face images generated by StyleGAN. A free edition with 10,000 128x128 synthetic images is made available for academic research. The website also provides an interactive face generator and an API Topics: ஏகத்துவத்தைக் கொண்டே பிரச்சாரத்தை ஆரம்பிக்க வேண்டும் and தாயத்து கட்டுவது ஷிர்க்கை சார்ந்தது Begin propagation with Monotheism, and Using amulets is Shirk Speaker: மவ்லவி கே.எல்.எம். generated.photos datasets 22 23 24 [ ]: இந்த தரவு அட்டவணையில் 19,457 முகமூடிகள் உள்ளன, இவற்றில், ஆன்லைனில் சேகரிக்கப்பட்ட 175 பொய்யான வீடியோக்களில் இருந்து உருவாக்கப்பட்ட 7,948 deepfake images மற்றும் பல்வேறு ஆன்லைன் தரவு அட்டவணையில் சேகரிக்கப்பட்ட 11,509 உண்மையான முகமூடிகள் உள்ளன. (Table 2 of the paper shows the dataset size is 19,509, but the dataset downloaded from pCloud contains only 19,457 images.) MesoNet Deepfake Dataset 1 [ ]: This dataset includes 100,000 synthesised face, bedroom, car and cat images by a GAN generator trained based on real images in the FFHQ and LSUN datasets (three object types – bedrooms, cars and cats – for the latter). Note that the name “100K-Generated-Images” was not a proper one as the authors [ ] just used this to name a sub-folder of their Google Drive shared space, but it was used in one of the survey papers [ ]. 100K-Generated-Images 30 25 26 30 65 [ ]: இந்த தரவு அட்டவணையில் 420,053 அற்புதங்கள் உள்ளன, இவற்றில் 156,930 Google Image API பயன்படுத்தி பதிவிறக்கப்பட்ட உண்மையான அட்டவணிகள் மற்றும் 263,123 இரண்டு வேறு வழிமுறைகளைப் பயன்படுத்தி உருவாக்கப்பட்ட பொய்யான முகமூடிகள் (Nirkin's Method மற்றும் Auto-Encoder-GAN) Ding et al.’s swapped face dataset 17 [ ]: This dataset includes 87,000 224x224 face images, generated by processing some StyleGAN-generated synthetic images using the GAN-fingerprint Removal approach (GANprintR) proposed by . It is the replaced version of the dataset, which contains 150,000 face images generated using an earlier version of GANprintR. iFakeFaceDB 48 Neves et al. FSRemovalDB [ ]: This dataset includes 40,000 images, half real and half deepfake. The images were collected from four sources: the CelebA-HQ dataset , the Flickr-Faces-HQ dataset , the 100K-Faces dataset (not available any longer, see the description of generated.photos datasets), and . Faces-HQ 21 27 28 29 thisperson-doesnotexist.com [ ]: This dataset includes 625,537 synthesised face images of 10,177 celebrities, with 43 rich attributes on face, illumination, environment and spoof types. The real images were selected from the CelebA dataset . The 43 attributes include 40 for real images, covering all facial components and accessories (e.g., skin, nose, eyes, eyebrows, lip, hair, hat, eyeglass), and 3 for fake images, covering spoof types, environments and illumination conditions. CelebA-Spoof 75 30 [ ]: This dataset contains 299,039 images, including 58,703 real images sampled from three datasets (FFHQ , CelebA and FaceForensics++ ) and 240,336 fake ones in four main facial manipulation types (identity swap, expression swap, attribute manipulation, and entire synthesis). The images cover two genders (male and female), a wide age groups (the majority between 21 and 50 years old), and both low- and high-quality levels. Diverse Fake Face Dataset (DFFD) 11 31 32 33 4.2 Deepfake Video Datasets [ ]: This dataset contains 620 deepfake face videos, generated by face swapping without manipulation of audio, covering 32 subjects and two quality levels (high and low). DeepfakeTIMIT 35 (FF) [ ]: This dataset contains 1,004 face videos with over 500,000 frames, covering various quality levels and two types of facial manipulation. This dataset is now replaced by the larger FaceForensics++ dataset (see below). FaceForensics 55 (FF++) [ ]: This dataset contains 5,000 face videos with over 1.8 million manipulated frames, including 1,000 real videos (with 509,914 frames) downloaded from YouTube, and 4,000 fake videos created using four face manipulation methods (Deepfakes, Face2Face, FaceSwap and NeuralTextures). The videos cover two genders (male and female), and three quality levels (VGA/480p, HD/720p, and FHD/1080p). FaceForensics++ 56 [ ]: This dataset contains 98 face videos, half (49) are real ones downloaded from Youtube, and the other half are fake ones generated using the FakeApp mobile application (which is now discontinued). The video dataset was created to used to demonstrate a deepfake video detection method based on detection of eye blinking behaviours, so all videos contain at least one eye-blinking event. All fake videos were created by swapping the original face in each of the real videos with the face of the actor Nicolas Cage , thus, only one subject is represented. UADFV dataset 39 34 [ ]: This dataset contains 142 “in the wild” deepfake portrait videos, collected from a range of online sources including news articles, online forums, mobile apps, and research presentations. The videos are diverse, covering the source generative model, resolution, compression, illumination, aspect-ratio, frame rate, motion, pose, cosmetics, occlusion, content, and context. Deep Fakes Dataset 10 [ ]: This dataset contains 5,244 face videos of 66 subjects with both face and voice manipulation. It was released as a preview of the full dataset of the 2020 Deepfake Detection Challenge (DFDC, see below). DFDC (Deepfake Detection Challenge) preview dataset 18 This data set contains 1,203 face videos of celebrities, including 408 real videos collected from YouTube with subjects of different ages, ethical groups and genders, and 795 deepfake videos synthesized from these real videos. Celeb-DF v1 35 [ ]: This dataset contains 6,229 face videos of celebrities, including 590 real videos collected from YouTube with subjects of different ages, ethic groups and genders, and 5,639 deepfake videos synthesised from these real videos. Celeb-DF v2 40 [ ]: This dataset contains 3,363 face videos, covering 28 subjects, gender, and skin colour. It was created as a joint effort between two units of Google, Inc.: Google AI and JigSaw . DeepFake Detection (DFD) Dataset 20 36 37 [ ]: This dataset contains 60,000 indoor face videos (with 17.6 million frames) generated by face swapping, covering 100 subjects, four skin tones (white, black, yellow, brown), two gen-ders (male and female), different age groups (20-45), 26 nationalities, 7 different angles, 8 face expressions, and different head poses. DeeperForensics-1.0 27 [ ]: This dataset contains 128,154 face videos of 960 subjects, including 23,654 real videos from 3,426 paid actors and 104,500 deepfake videos created using eight different methods (DF-128, DF-256, MM/NN face swap, NTH, FSGAN, StyleGAN, refinement, and audio swap). DFDC (Deepfake Detection Challenge) full dataset 18 10 [ ]: This dataset contains 10,000 high-quality forgery videos, with video- and face-level annotations. The dataset focuses on a more challenging case for forgery detection: each video involves one to 15 individuals, but only some (a minority of) faces are manipulated. FFIW K (Face Forensics in the Wild) dataset 79 FFIW [ ]: This dataset contains 37,942 videos of paid subjects (395 Koreans and 8 Southeastern Asians), including 62,166 real videos and 175,776 fake ones created using six methods – FaceSwap, DeepFaceLab, FSGAN, First Order Motion Model (FOMM), Audio-driven Talking Face HeadPose (ATFHP) and Wav2Lip. The videos cover a balanced gender ratio and a wide range of age groups. Korean DeepFake Detection Dataset (KoDF) 36 [ ]: This dataset contains 1,737 videos with 1,666,816 frames, including 1,339,843 real frames and 326,973 fake frames generated using the Deep Video Portraits (DVP) [ ] method. The original videos were obtained from three sources: the dataset used in [ ], the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) [ ], and YouTube. Most videos have a resolution of 1280×720. VideoForensicsHQ 23 34 33 42 [ ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] WildDeepfake 81 4.3 Deepfake Audio/Speech Datasets Voice conversion (VC) is a technology that can be used to modify an audio and speech sample so that it appears as if spoken by a different (target) person than the original (source) speaker. Obviously, it can be used to generate deepfake audio/speech samples. The biennial Voice Conversion Challenge that started in 2016 is a major challenge series on VC. Datasets released from this challenge series are very different from other deepfake datasets: the deepfake data is not included in the original dataset created by the organisers of each challenge, but in the participant submissions (which are retargeted/fake utterances produced by VC systems built by participants). The challenge datasets also include the evaluation (listening-based) results of all submissions. Some fake utterances may be produced by DL-based VC systems, so we consider all datasets from this challenge series relevant for our purpose of this survey. 38 [அரசன் ]: The original dataset created by the challenge organisers was derived from the DAPS (Device and Produced Speech) Dataset [ ]. It contains 216 utterances (162 for training and 54 for testing) per speaker from 10 speakers. Participating teams (17) developed their own VC systems for all 25 source-target speaker pairs, and then submitted generated utterances for evaluation. At least six participating teams used DL-related techniques (LSTM, DNN) in their VC systems (see Table 2 of the result analysis paper ), so the submitted utterances can certainly be considered deepfakes. Voice Conversion Challenge 2016 database 62 47 39 [அரசன் ]: The original dataset created by the challenge organisers was also based on the DAPS dataset. It contains 116 utterances (81 for training and 35 for testing) per speaker from 12 speakers in two different tasks (called Hub and Spoke). Participating teams (23 in total, all for Hub and 11 for Spoke) developed their own VC systems for all 16 source-target speaker pairs, and then submitted generated utterances for evaluation. Comparing with the 2016 challenge, more participating teams used DL-related techniques (e.g., WaveNet, LSTM, DNN, CycleGAN, DRM – deep relational models, and ARBM – adaptive restricted Boltzmann machines) in their VC systems. Voice Conversion Challenge 2018 database 44 [அரசன் ]: This dataset is based on the Effective Multilingual Interaction in Mobile Environments (EMIME) dataset , a bilingual (Finnish/English, German/English, and Mandarin/English) database. It contains 145 utterances (120 for training and 25 for testing) per speaker from 14 speakers for two different tasks (with 4 × 4 and 4 × 6 source-target speaker pairs, respectively). Participating teams (33 in total, out of which 31 for Task 1 and 28 for Task 2) developed their own VC systems for all source-target speaker pairs, and then submitted generated utterances for evaluation. Comparing with the 2018 challenge, DL-based VC systems were overwhelmingly used by almost all participating teams (WaveNet and WaveGAN among the most used DL-based building blocks). Voice Conversion Challenge 2020 database 70 40 A major set of deepfake speech datasets were created for the (Automatic Speaker Verification Spoofing and Countermeasures) Challenge (2015-2021, held biannually). The datasets for the 2019 and 2021 contain speech data that can be considered deepfakes. ASVspoof 41 [அரசன் ]: This dataset is based on the Voice Cloning Toolkit (VCTK) corpus , a multi-speaker English speech database captured from 107 speakers (46 males and 61 females). Two attack scenarios were considered: logical access (LA) involving spoofed (synthetic or converted) speech, and physical access (PA) involving replay attacks of previously recorded bona fide recordings). For our purpose in this survey, the LA scenario is more relevant. The LA part of the dataset includes 12,483 bona fide (real) utterances and 108,978 spoofed utterances. Some of the spoofed speech data for the LA scenario were produced using a generative model involving DL-based techniques such as long short-term memory (LSTM) , WaveNet [ ], WaveRNN [ ], WaveCycleGAN2 [ ]. Note that the challenge organisers did not use the term “deepfake” explicitly, despite the fact that the DL-generated spoofed speech data can be considered as deepfakes. ASVspoof 2019 Challenge database 67 42 43 50 28 58 [அரசன் ]: This dataset contains bona fide and spoofed speech data for the logical access (LA) task. The challenge is still ongoing and we did not find a detailed paper on the dataset, so cannot include more details other than its size (7.8 GB after compression). Although we did not see details of the generative algorithms used to produce spoofed speech data, we believe similar DL-based algorithms were used like for the 2019 challenge. ASVspoof 2021 Challenge – Logical Access Database 14 [ ]: In 2021, the challenge included an explicitly defined track on deepfake, but the task description suggests that the organisers of the challenge considered a broader definition of the term “deepfake” by looking at spoofing human listeners rather than ASV (Automatic Speaker Verification) systems. The size of the dataset is 34.5 GB after compression. ASVspoof 2021 Challenge – Speech Deepfake Database 15 Possibly because of the long history and wide participation of the community in the ASVspoof challenges for creating the dedicated datasets, there are very few other deepfake audio/speech datasets. One such dataset was created by a group of researcher from Baidu Research [ ]. This dataset was created to demonstrate a proposed voice cloning method. It is relatively small, and contains 134 utterances, including 10 real ones, 120 cloned ones, and 4 manipulated ones. Another dataset was created by Google AI and Google News Initiative , but it was made part of the ASVspoof 2019 dataset. This dataset contains thousands of phrases spoken by 68 synthetic “voices” covering a variety of regional accents. 5 44 4.4 Hybrid Deepfake Datasets : These datasets were created by the DARPA Media Forensics (MediFor) Program 2020 இல் OpenMFC . There are two GAN-generated deepfake datasets, one with more than 1,000 deepfake images and the other with over 100 deepfake videos. The datasets were made available to registered participants of the competition only. NIST OpenMFC (Open Media Forensics Challenge) Datasets 45 46 47 [ ]: This dataset is named as “a versatile benchmark for comprehensive forgery analysis”. It contains 2,896,062 images and 221,247 videos, including 1,457,861 fake images and 121,617 fake videos. The videos and images cover seven image-level and eight video-level manipulation approaches, 36 different types of perturbations and more mixed perturbations, and a large number of annotation labels (6.3 million classification labels, 2.9 million manipulated area annotations and 221,247 temporal forgery segment labels). The dataset is being used for supporting the Face Forgery Analysis Challenge 2021 at the SenseHuman 2021 (3rd Workshop on Sensing, Understanding and Synthesizing Humans) ICCV 2021 இல் நடந்த ஆய்வு . ForgeryNet 25 48 49 50 4.5 A Deepfake Dataset Generator [ ]: This is not actually a dataset per se, but a system for producing large datasets more automatically, including generating deepfake datasets. One may argue the automatically generated datasets are fake since they are not produced from real-world scenes. DatasetGAN 74 4.6 Different Databases இல் Deepfakes பற்றிய Subjective Quality பிரிவில் கூறியது போல , subjective quality evaluation is necessary to evaluate the realness, realisticness, and naturalness of deepfake media. While there has been very limited work on this topic, in 2020, Jiang et al. [ ] conducted a user study on realness of deepfake videos. They recruited 100 professional participants (most of whom are computer vision researchers), who were asked to evaluate the realness of 30 randomly selected videos from 7 deepfake video datasets (DeeperForensics-1.0, UADFV, DeepFake-TIMIT, Celeb-DF, FaceForensics++, Deep Fake Detection, and DFDC). Participants were asked to respond to the statement “The video clip looks real.” and gave scores following a five-point Likert scale (1 – clearly disagree, 2 – weakly disagree, 3 – borderline, 4 – weakly agree, 5 – clearly agree). 4.7 27 அலுவலகம் shows the results. Interestingly, we can see a huge difference between the realness levels of different datasets. What is probably quite surprising is that FaceForensics++, one of the most widely used deepfake datasets, has a very low MOS score and less than 9% of participants considered the 30 selected videos as real. 3 Table 3: Human-judged subjective quality (realness) of deepfake videos in 7 datasets. The MOS scores were not reported by , but calculated by us based on the raw data shown in Table 3 of [ ]. Jiang et al. 27 4.7 Discussion: Datasets Among all deepfake image and video datasets, a significant majority are about face images and videos. This is not surprising since face swapping, face attribution manipulation, and fully synthesised face images are among the hottest topics within deepfake research and real-world applications. We hope more non-face deepfake image and video datasets can be produced to support a broader range of research activities on deepfake. The subjective quality results shown in Table indicate that it is important to check realness of deep-fake media to support any performance evaluation or comparison. To ensure that the quality evaluation of datasets is fair, transparent and reliable, standard procedures need defining and a common pool of qualified human experts should be used. 3 Many authors of deepfake-related datasets attempted to classify such datasets into different generations. Chronologically speaking, we could broadly split such datasets into two generations: before 2019 and since 2019. Typically, datasets created before 2019 are relatively less advanced and smaller, while those created after 2019 tend to be larger, more diverse (i.e., covering more attributes), and of higher quality (i.e., produced by more advanced generative models). This can also be seen from the data in Table , in which the top two datasets (DeeperForensics-1 and Celeb-DF) fall within the new generation (2020), while others belong to the old generation. In addition to the two generations, a newer generation has also emerged in 2021: a number of very recent datasets started focusing on more realistic deepfakes (i.e., in the wild) or more specified areas of deepfakes (e.g., 10 focusing on multiple faces in the same video, and KoDF focusing on Korean faces). This trend shows that the deepfake research community has grown significantly in the past few years so that narrower topics have also started gaining attention and interest from some researchers. 3 FFIW K 5 Deepfake-Related Challenges, Competitions & Benchmarks (இந்திய மொழிபெயர்ப்பு) This section reviews initiatives aiming to advance the state-of-the-art of detection and generation of synthetic or manipulated media (such as video, image and audio) via competitions or challenges open to the public, and ongoing benchmarks tackling specific problems. 5.1 Manipulated Media கண்டுபிடிப்பு The Deepfake Detection Challenge (DFDC) இது ஒரு AI and Media Steering Committee-ல் உருவாக்கப்பட்டது. , including BBC, Facebook, Amazon, Microsoft and New York Times, and some universities around the world including the University of Oxford. The competition remained open from 5 September 2019 till 31 March 2020, and involved 3 stages. At first, the DFDC preview dataset was released. At a later stage, the DFDC full dataset was also made available to the 2,114 participants of the competition incorporating face and audio swap techniques for generation of deepfake content. At the final stage, the submitted models were evaluated using a test dataset (referred to as the “black box dataset”) of 10,000 videos which included deepfake videos. The best performance on the black box dataset had an accuracy of 65.18%, according to the released results [ ]. Submissions were ranked according to the overall log loss score, as defined in Eq. ( ). All top five ranked models (the winner had the lowest overall log loss) are available on GitHub. Results indicate how challenging the detection of deepfake is since the best accuracy was low and “ ”, according to Dolhansky et al. [ ]. Figure shows a screenshot of the leaderboard with the five finalists. The first top ranked model used MTCNN (Multi-tasked Cascaded Convolutional Network), the second used WS-DAN (Weakly Supervised Data Augumentation Network), and the third used the EfficientNetB7 architecture. Meta compiling the common themes observed in the winning models, they were: clever augmentations, architectures, and absence of forensics methods. Moving forward, they called for “ » 51 52 in-the-wild 22 53 13 many submissions were simply random 19 2 solutions that go beyond analysing images and video. Considering context, provenance, and other signals may be the way to improve deepfake detection models The Automatic Speaker Verification Spoofing And Countermeasures Challenge Workshop (ASVspoof) has been running biennially since 2015. This competition is organised by an international consortium that includes Inria and EURECOM (France), University of Eastern Finland, National Institute of Informatics (Japan), and Institute for Infocomm Research (Singapore). This year the ASVspoof challenge includes, for the first time, a sub-challenge focused on where the envisioned use case is an adversary trying to fool a human listener. The metric used for evaluating performance of submitted solutions (i.e., classifiers) is EER. Four baseline solutions (also called “countermeasures”), each using a different technique, were made available to participants with their corresponding EER metric values. The ASVspoof 2021 Speech Deepfake Database containing audio recordings with original and spoofed utterances has also been made available. The competition involves three phases : a progress phase, an evaluation phase and a post-evaluation phase; it is unclear how teams move from one phase to the next. More information about the 2021 competition is available in the published evaluation plan [ ]. The organisers of the competition noted that they opted for the EER as the performance evaluation met-ric for countermeasures submitted to the speech deepfake task for legacy reasons. They acknowledged, however, that “ ” by the ISO/IEC 19795-1:2021 standard. Despite the fact that only the 2021 ASVspoof competition contained a track explicitly related to deepfake, some data in the ASVspoof 2019 dataset (Logical Access task) used for the 2019 competition was generated using DL-based algorithms as mentioned in Section . We expect that this also holds for the ASVspoof 2021 dataset (Logical Access task). The ASVspoof 2019 competition used the EER as secondary metric; the primary performance metric used was the tandem detection cost function (t-DCF) [ ]. According to its evaluation plan [ ], t-DCF assesses the performance of the whole tandem system whereby “ ”. It is calculated according to Eq. ( ), where cm ( ) and cm( ) are, respectively, “ ”. 54 Speech DeepFake 55 56 13 EER reporting is deprecated 57 4 63 69 a CM [countermeasure] serves as a ‘gate’ to determine whether a given speech input originates from a bona fide (genuine) user, before passing it the main biometric verifier (the ASV system) 17 P s P s the miss rate and the false alarm rate of the CM system at threshold s For further information about Eq. ( ), including constants 1 and 2, please refer to the ASVspoof 2019 evaluation plan [ ]. 17 C C 69 An implementation of the t-DCF metric has been made available by the ASVspoof 2019’s organisers in Python and Matlab formats. 58 59 The Face Anti-spoofing (Presentation Attack Detection) Challenge started in 2019. Its first two editions were held at the 2019 and 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2020), respectively. Its third edition was moved to be co-located with the 2021 IEEE/CVF International Conference on Computer Vision (ICCV 2021). This competition series was organised by a group of researchers from academia and industry in China, Mexico, Spain, Finland and the US. The 2021 competition was focused on 3D high-fidelity mask attacks, and followed a 2-phased process. The first phase is the “development phase”; it started in April 2021 when the CASIA-SURF HiFiMask dataset was released to participants. The second phase is the “final ranking phase” (June 2021), when the competition ended. The competition adopted the following performance metrics for evaluation of the solutions submitted: attack presentation classification error rate (APCER), normal/bona fide presentation classification error rate (NPCER/BPCER), and average classification error rate (ACER), in accordance with the ISO/IEC 30107-3:2017 standard. Figure provides the leaderboard for the top three solutions. 60 61 62 63 64 3 The FaceForensics Benchmark is an ongoing automated benchmark for detection of face manipulation. The organisers of the benchmark made the FaceForensics++ dataset available for training. Manipulated videos (4,000 in total) were created using four techniques, i.e., two computer graphics-based approaches (Face2Face and FaceSwap) and two learning-based approaches (DeepFakes and Neural Textures). The deepfakes videos were generated using a slightly modified version of FaceSwap , and the Neural Textures videos were created using the approach proposed by Thies et al. [ ]. The benchmark test dataset is created from the collection of 1,000 images randomly selected from either the manipulation methods or the original videos [ ]. Participants have to submit results to the benchmark, rather then code like other competitions; this is illustrated in Figure . The outcome of a submission is illustrated in Figure , where the scores are a measure of accuracy (Eq. ( )). 65 66 61 56 4a 4b 8 The Open Media Forensics Challenge (OpenMFC, formerly DARPA MFC) is an annual image and video forensics evaluation aiming to facilitate development of multimedia manipulation detection systems. It has been organised annually starting from 2017 under the name of DARPA MFC. In 2020, the National Institute of Standards and Technology (NIST) initiated the as a new evaluation platform, based on their previous experiences with the DARPA MFC series, to make the participation more convenient for all researchers. In OpenMFC 2020, two deepfake-related tasks were included for the first time: Image GAN Manipulation Detection (IGMD) and Video GAN Manipulation Detection (VGMD). The organisers provided an image evaluation dataset for the IGMD task, containing 1,000 images from over 200 image journals , and a video evaluation dataset for the VGMD task, including over 100 test videos. used in the previous MFC challenges as development datasets. The challenge is composed of two main phases for development and evaluation, respectively, and a pre-challenge phase for quality control testing. For evaluation of submissions, AUC-ROC is used as the primary metric. Furthermore, CDR@FAR, where CDR refers to correct detection rate or TPR (Eq. ( )) and FAR refers to false alarm rate or FPR (Eq. ( )), is also used as a metric [ ]. The DeeperForensics Challenge 2020 is a deepfake face detection challenge held at the 2020 ECCV 67 68 OpenMFC 69 70 4 5 49 71 SenseHuman Workshop . The challenge used the DeeperForensics1.0 dataset. 72 The organisers provided a hidden test dataset to better simulate real-world scenarios. The challenge involved two phases: the “development phase” that started in August 2020 allowing 100 successful sub-missions, and the “final test phase” that started in October 2020 allowing 2 successful submissions until the end of the month. The submissions were evaluated using the binary cross-entropy loss (BCELoss) metric, calculated according to Eq. ( எங்கே is the number of videos in the hidden test set, is the ground truth label of video (fake:1, real:0), and ( ) is the predicted probability that video is fake. 18 N yi i p yi i Results of the competition were discussed by Jiang et al. [ ]. The top solution used three models, i.e., EfficientNet-B0, EfficientNet-B1 and EfficientNet-B2, for classification. The second top used EfficientNet-B5 for both an image-based model and a video-based model. The third ranked solution used a 3D convolutional neural network (3DCNN). 73 26 The Face Forgery Analysis Challenge 2021 is a competition hosted at the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2021). It is organised by researchers from a number of organisations in China including universities and SenseTime Research (the research arm of SenseTime , one of the major AI “unicorns” in China). The challenge aims to advance the state-of-the-art in detection of photo-realistic manipulation of images and videos. Participants are able to use a large annotated face dataset (i.e., the ForgeryNet dataset) that was obtained by applying a number of techniques for manipulation (15) and perturbation (36) to train their solutions. The phases comprise of Forgery Image Analysis, Forgery Video Analysis, Forgery Video Temporal Localization phases, and the final phase (i.e., “private test”) where participants’ models will be tested against an unseen dataset. The following metrics will be used [ ]: AUC, average precision (AP) at some “temporal Intersection over Union” (AP@tIoU) compared to a threshold ∈ [0*. ,* 0*.*95], and average recall (AR) at (AR@ ) where is the top labels returned for multi-class classifiers. 74 75 25 t 5 K K K K The 2020 CelebA-Spoof Face Anti-Spoofing Challenge was hosted at the 16 European Conference on Computer Vision (ECCV 2020). The challenge ran between August and October 2020, and aimed to advance the state-of-the-art in detecting “ ” [ ]. The organisers made the face CelebA-Spoof dataset available for the competition containing rich annotation across a range of attributes. The competition only had one phase where participants submitted their solutions to be evaluated against a test dataset; the spoof class was considered as “positive” and the live class as “negative”. Metric TPR@FPR was used and collected at three points where the TPR when FPR = 10 4 determined the final ranking. The top three finalists (see Figure ) used deep learning models ResNet, EfficientNet-B7, and a novel architecture combining Central Difference Convolutional Networks (CDCN) and Dual Attention Network (DAN). The two top ranked solutions used different strategies to boost their models’ performance: a heuristic voting scheme was used by the top-ranked solution, and a weight-after-sorting strategy was used by the second ranked solution. 76 th whether a presented face is live or spoof 76 − 5 The 2021 CSIG Challenge is the second edition of a challenge organised by the China Society of Image and Graphics . The 2021 challenge has the Fake Media Forensic Challenge as its 6 track, co-organised by CSIG’s Digital Media Forensics and Security Technical Committee and Institute of Information Engineering, Chinese Academy of Sciences . This track has two tasks, one on deepfake video detection, and the other on deepfake audio/speech detection. For the deepfake video detection task, the dataset used contains a public training set with 10,000 sound-free face videos (including 4,000 fake videos), a public test set with 20,000 face videos (the percentage of deepfake videos is unknown to participants), and a private test set that will be determined and used at the final session for selecting the winners. All videos contain faces of Eastern Asian people, and cover a wide range of parameters such as multiple resolutions and encoding quality factors, the use of blurring or sharpening filters, and added noise. Deepfake videos were created using public tools including DeepFaceLab [ ], Faceswap , Faceswap-GAN, Recycle-GAN [ ] and ALAE (Adversarial Latent Autoencoders) [ ]. For the deepfake audio/speech detection task, the dataset used contains a public training set with 10,000 speech samples (including 6,000 fake ones), a public test set with 20,000 face videos (the percentage of deepfake videos is unknown to participants), and a private test set for the final session (the same as the deepfake video detection task). The tools used for generating the fake speech samples include TTS (text-to-speech) voice synthesis tools and VC (voice conversion) tools. The main TTS tools used include open-source tools such as DeepVoice, TensorFlowTTS and GAN-TTS [ ] and commercial software tools such as those from iFlytek போன்ற வணிக மென்பொருள் கருவிகள் and IBM. The main VC tools used include Adaptive-VC and CycleGAN-VC [ ]. For both deepfake detection tasks, the performance metric used is log loss. 77 78 79 th 80 81 53 82 6 54 83 8 84 29 2020 China Artificial Intelligence was the second edition of a Chinese AI competition open for the general public to participate, organised by the municipal government of the City of Xiamen in China. In 2020, it had two sub-competitions, Multimedia Information Recognition Technology Competition and Language and Knowledge Technology Competition . The Multimedia Information Recognition Technology Competition included two tasks on deepfakes: one on deepfake video detection and one on deepfake audio/speech detection . The deepfake video detection task used 3,000 videos, and log loss was used as the sole performance metric. The deepfake audio/speech detection task used 20,000 audio samples (mostly in Chinese, and the remaining in English), and EER was used as the sole performance metric. For both tasks, the ratio between real and deepfake samples was 1:1. We did not find where to download the datasets used for the tasks nor a more detailed technical description of the datasets. For the deepfake video detection tasks, the top two winning teams (with an A prize) were from Netease (Hangzhou) Network Co., Ltd. and Beijing RealAI Technology Co., Ltd., followed by three other teams winning a B prize: Xiamen Fuyun Information Technology Co., Ltd.; Institute of Computing Technology, Chinese Academy of Sciences; and Wuhan Daqian Information Technology Co., Ltd. For the deepfake audio/speech task, there was no team winning an A prize, but one team winning a B prize: SpeakIn Technologies Co., Ltd. The final results of some teams were published, but some teams were allowed to hide their results. We did not find a detailed technical report summarising the results and explaining the work of the winning teams. 85 86 87 88 89 B-prize வெற்றியாளர் அணி ஒன்றாகும் Beijing RealAI Technology Co., Ltd., deepfake-related R&D இல் செயல்படும் ஒரு சீனா நிறுவனமாகும். 5.2 Generation of Manipulated Media The Voice Conversion Challenge is a biennial competition that has been running since 2016. The challenge and the corresponding workshop, hosted at the INTERSPEECH conference , is supported by the SynSig (Speech Synthesis Special Interest Group) International Speech Communication Association (ISCA) என்ற பிரிவில் . Its aim is to promote progress in voice conversion (VC) technology that can be applied to a number of positive and negative use cases, such as spoofing voice biometric systems. The 2020 challenge focused on speaker conversion, a sub-problem of VC, and included two tasks. For the first task “intra-lingual semi-parallel voice conversion”, participants had to develop 16 VC systems (speaker-pair combinations) including male and female speakers and English sentences, using the provided Voice Conversion Challenge 2020 database v1.0 for training (refer to Section ). For the second task “cross-lingual voice conversion”, participants had to develop 24 VC systems, also including male and female speakers, but uttering sentences in three languages (Finnish, German and Mandarin), based on the provided training dataset. Figure illustrates the process of training and generation of VC systems. 90 91 92 93 4 6 Submissions were evaluated for “ ” . As such, the organisers used [ ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] perceived naturalness and similarity through listening tests 94 subjective evaluation 70 The Deepfake Africa Challenge (2021) is a new initiative of the AI Africa Expo, in partnership with a film and media production company (Wesgro) and the African Data Science competition platform Zindi. Its aim is “ ”. Eligible participants were required to be citizens and residents of the African continent. Submissions, accepted up to end of July 2021, can be either video or audio. Evaluation of submissions is defined in terms of artistic creativity, relevance of challenge topic, and innovation in the process of generation as long as participants use tools and packages publicly available. The top three finalists will receive a prize, present their work at the Expo, and will have to grant copyrights to Zindi. Unlike the other competitions reviewed in this section, which were focused on advancing the state-of-the-art in detection of synthetic or manipulated media, this competition focused on the generation of deepfake which seems more humanities-centred. This is a trend observed in arts [ ] and culture [ ].5.3 Generation and Detection of Manipulated Media 95 to create convincing deepfakes to highlight the power of this synthetic media, illustrating its creative potential for exploitation for both positive and negative outcomes and focusing debate about its ethical use / misuse in an African context 31 57 The DeepFake Game Competition (DFGC) is in its first edition, hosted at the 2021 International Joint Conference on Biometrics (IJCB 2021). Its organisers are mainly from the Institute of Automation Chinese Academy of Sciences (CASIA). The idea of the competition was to promote an adversarial game between agents pushing for advances in both deepfake creation and detection. In order to achieve this, a 6-stage protocol was designed interleaving three creation phase (C-phase) and detection phase (D-phase), typically one week apart; submissions closed in April 2021. Both C-phases and D-phases were bound to the Celeb-DF (v2) dataset [ இதில், 6229 வீடியோக்கள் (590 உண்மையான / ஆரம்ப வீடியோக்கள் மற்றும் 5,639 பொய்யான / மாற்றப்பட்ட வீடியோக்கள்) உள்ளடக்கியுள்ளன, பயிற்சியாளர்களுக்கு பயிற்சியாளர்களுக்கு பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியாளர்கள் பயிற்சியா ]. The metrics used for evaluation were: a detection score, used for evaluation of a D-phase, and a creation score, used for evaluation of a C-phase. The top three finalists for the detection phase employed CNN-based classifiers EfficientNet-B3, Efficientnet-B0 and EfficientNetV2. 96 40 52 97 The Detection Score ( ) metric captures the models’ ability to correctly classify fake images submitted to the previous C-phase against a set of real images in the CelebDF test dataset. It is calculated using Eq. ( ), where is the number of valid submissions of created synthesis test sets in the last C-phase. டே 21 NC The Creation Score ( இந்நிகழ்ச்சியில் ஈடுபட்டு வருபவர்களின் எண்ணிக்கை அதிகரித்து வருகிறது. ), where is the number of valid submissions of detection methods in the last D-phase, the noise score ( noise) penalises noisy images, the other three parts of the equation relate to the following : “ ”. CS 22 ND S 98 ID level similarity to the donor ID, image level similarity to the target frame, and the deception ability against detection models. ID level similarity is scored by a face recognition model using dot product of two ID features (fake face ID and donor ID). The image level similarity is scored by SSIM [Structural Similarity Index] to make sure the face-swapped image is similar to the corresponding target image in content and quality பௌர்ணமி et al. [ ] observed a commonality between the three winning teams for the creation task, i.e., the use of the FaceShifter [ ] framework for face swapping. They highlighted two overall reflections about the competition: (1) the limited diversity of the deepfake datasets submitted and the use of repetitive methods to generate them, and (2) the limited size of the Celeb-DF (v2) dataset itself flagging the need for a larger dataset for next year’s competition. The organisers of the competition also applied the top two detection models to unseen datasets (DFDC and FaceForensics++) and noticed that they do not generalise well. 52 37 6 A Meta-Review of Deepfake-Related Surveys This section presents a meta-review of 12 selected deepfake-related survey papers, including eight published in English [ , , , – , , ] and four published in Chinese [ , , , ]. It covers the following aspects in a systematic manner: definitions and scope, performance metrics, datasets, challenges/competitions/benchmarks, performance comparison, key challenges and recommendations. 16 45 46 64 66 71 73 7 38 41 59 The meta-review aims at drawing some high-level insights for monitoring future development of deepfake-related technologies and their applications. 6.1 Definitions and Scope நாங்கள் பிரிவில் கூறியது போல , among researchers, practitioners and law makers there is no universally accepted definition of “deepfake” as a term. This is also reflected in how the authors of the 12 survey papers considered this aspect. Most authors talked about the history of deepfakes and pointed out that the term reflects the combination of “deep learning” and “fake”, but some used a broader definition, e.g., Lyu [ ] defined deepfake as “ ”. Some authors also referred to deepfake-related legislations, but none of them pointed out that the definitions in some such legislations are completely different from the more technical definitions involving the use of deep learning. No authors discussed the blurred boundary between deepfakes and non-deepfakes, although some surveys actually cover both, e.g., Tao et al. [ ] focused on speech forgery and did not explicitly highlight “deepfake”. 1.1 45 high quality fake videos and audios generated by AI algorithms 59 In terms of the scope, while some authors (correctly) considered all types of media that can be produced by deepfake-related techniques [ , , , இவ்வாறு ஒரு சில எழுத்தாளர்கள் கூறுகின்றனர். , , , இவ்வாறு எழுதியவர்களும், எழுத்தாளர்களும் மட்டுமே எழுதியிருக்கிறார்கள். , ] have considered images and videos. Another phenomenon we observed is that many authors focused more on face images and videos, and authors of three surveys [ , , ] “Deepfake” என்பதன் அர்த்தத்தை இவ்வளவு கடுமையான அளவுக்கு குறைக்கலாம்: 38 41 45 65 7 64 71 73 16 66 16 64 71 Deshmukh and Wankhade [ ] defined it as “ ”; 16 a technology which creates fake images or videos of targeted humans by swapping their faces [by] another character saying or doing things that are not absolutely done by them and humans start believing in such fake as it is not always recognisable with the everyday human eye Younus and Hasan [ ] considered deepfake as a technique allowing “ ”; and 71 any computer user to exchange the face of one person with another digitally in any video Tolosana et al. [ ] defined it as “ ”. 64 a deep learning based technique able to create fake videos by swapping the face of a person by the face of another person Such unnecessarily narrow definitions and scopes can lead to confusion and do not help exchanges between researchers and practitioners working on different types of deepfakes. We call on more researchers to accept a broader definition of “deepfake” so that highly realistic/natural media of any kind generated by a sophisticated automated method (often AI-based) is considered deepfake. Here, we provide two examples of such a broader definition: the image2image (or pixel2pixel) technique [ ] that allows the production of deepfake images and videos of any objects (e.g., the “horse2zebra” deepfake image shown in Figure ), and the the so-called “deepfake geography [ ]”, where AI-based techniques are used to generate realistic-looking satellite images. 80 7 77 Another important fact missed or not sufficiently discussed by authors of all the 12 surveys is that deepfake techniques can be used for positive applications, e.g., creative arts, entertainment and protecting online users’ privacy. We call for more researchers and practitioners to follow the proposal in the 2020 Tencent AI White Paper [ ] to start using the more neutral-sounding term “deep synthesis”. Accordingly,we can use different words for different types of data generated using “deep synthesis” techniques, e.g., “deep art”, “deep animation”, “deep music”, and “deepfake”. While authors of the 12 survey papers did not recognise the positive applications of “deepfake” technologies, some other researchers did, e.g., organisers of the Voice Conversion Challenge 2020 who said the VC technology (for speech deepfake) “ ”. 60 99 is useful in many applications, such as customizing audio book and avatar voices, dubbing, movie industry, teleconferencing, singing voice modification, voice restoration after surgery, and cloning of voices of historical persons 6.2 Performance Metrics ஒரு முதியவன் பாதாளங்களைத் தாண்டும் தன் மந்திரக்கோலால் சாய்த்தபடியிருக்கிறான் நாட்சத்திரங்களை. .............................................................................................................................................................................. இது எத்தனையாவது [...] 6.3 Datasets Many of the 12 survey papers list a number of deepfake-related datasets, but none of them have coverage as complete as ours shown in Section உதாரணமாக, எந்த ஆய்வுகளிலும் Voice Conversion Challenge 2016/2018/2020 தரவு அட்டவணைகள் கையாளப்படவில்லை மற்றும் ASVspoof 2019/2021 தரவு அட்டவணைகள் இரண்டு ஆய்வுகளில் மட்டுமே சிறிது நேரம் கையாளப்பட்டுள்ளன. , ]. In addition, more recent deepfake datasets especially those released in 2021 are also not covered by any of the surveys. We believe that our Section is the most comprehensive review of deepfake-related datasets so far. 4 38 59 4 Some survey papers include datasets that are likely deepfakes, e.g., Verdoliva [ ] covered many general fake image datasets where the manipulated images were not generated by deep learning or even AI-based methods, and some surveys (e.g., [ ]) mentioned ASVspoof 2015 datasets but we did not see the use of deep learning for generating data used in the dataset. 66 38 6.4 Challenges, Competitions and Benchmarks Many surveys cover deepfake-related challenges, competitions and benchmarks. The coverage is, however, mostly limited, and some challenges (e.g., the Voice Conversion Challenge 2016/2018/2020 and the two Chinese challenges we covered in Section ) are not covered by any of the surveys. The level of detail of challenges, competitions and benchmarks is also normally limited, compared with what we chose to include in Section . Similar to the datasets we covered in Section , we believe that our coverage of deepfake-related challenges, competitions and benchmarks in Section is also the most comprehensive so far. 5 5 4 5 6.5 Performance Comparison Most surveys have a good coverage of related methods for deepfake generation and detection, but only some explicitly covered performance comparison between different methods [ , , ]. 38 46 64 Among all the survey papers, Li et al. [ ] conducted the most comprehensive study on performance of different deepfake detection methods. In addition to showing the performance metrics of a number of deepfake detection methods in Table 3 of [ ], they also looked at general characteristics and issues of different types of deepfake detection methods, as shown in Table . Furthermore, they also looked at research on robustness of deepfake detection methods against adversarial samples, referring to some work that showed a lack of such robustness. 38 38 4 Due to quality issues of many deepfake-related datasets (discussed in Section ), we need to treat any performance metrics and comparison of different detection methods with caution. Without testing all methods on a sufficiently large, diverse and high-quality deepfake dataset, the performance comparison results can be misleading. This highlights the importance of having more challenges, competitions and benchmarks to encourage performance comparison on standard datasets and using consistent performance metrics. 4.6 6.6 Challenges and Recommendations The authors of some surveys identified some key challenges and future research directions for the deepfake community. Not surprisingly, how to develop more robust, scalable, generalisable and explainable deepfake detection methods is one of the most discussed key challenges and also a major future research direction [ , , , , , , , , ]. Considering the arms race between deepfake generation and detection, this research direction will likely remain the hottest topic in deepfake research. 7 16 38 41 45 59 65 66 71 A couple of surveys [ , ] mentioned fusion as a key future research direction, where “fusion” refers to combining different methods (e.g., combining multiple detectors of different types) and data sources (e.g., jointly considering audio-visual analysis) to achieve better performance for deepfake detection. Lyu [ ] suggested that, for detection of deepfake videos, we need to consider video-level detection more, which can be considered fusion of detection results of all video frames. 38 66 45 The authors of three surveys, Lyu [ ] , Deshmukh and Wankhade [ ] and Younus and Hasan [ ], argued that better (higher-quality, more up-to-date, and more standard) deepfake datasets are needed to develop more effective deepfake detection methods. Lyu [ ] also suggested that we need to consider effects in training data and improve the evaluation of datasets. We agree with them on these points. 45 16 71 45 social media laundering Tao et al. [ ] suggested that low-cost deepfake generation/detection should be considered as a future research direction. This is a valid recommendation since lightweight methods will allow less powerful computing devices (e.g., IoT devices) to benefit from such technologies. 59 Two Chinese surveys [ , ] also mentioned the need to have new deepfake-related legislations combating malicious use of deepfakes and the need to train end users such as journalists. This is likely an area where interdisciplinary research can grow. 38 41 There are also other ad-hoc recommendations given by the authors of some surveys. For example, Lyu [ இவ்வாறு கூறுகையில், பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வகையான, பல்வேறு வ ] discussed specific research directions for different deep-fake generation methods (face synthesis, identity swap, attribute manipulation, and expression swap). Liang et al. [ ] and Li et al. [ ] recommended more active defence mechanisms such as using digital watermarking and blockchain technologies to build trustworthy media frameworks against deepfakes. 45 64 41 38 7 Conclusion The rapid growth in the capability to manipulate media or create synthetic media which look realistic and natural paved the way for deepfakes. At first, this paper adopted a critical approach to look at different definitions of the term “deepfake”. In that regard, we point out the different contradicting definitions and call for the wider community to consider how to define a new term that has a more consistent scope and meaning. For instance, replacing “deepfake” by “deep synthesis” can be more inclusive by embracing positive applications of deepfake techniques, e.g., in entertainment and for simulation purposes. This paper provided a comprehensive overview of multiple aspects of the deepfake ecosystem drawing from the research literature and other online sources published in two languages: English and Chinese. It covers commonly used performance metrics and standards, related datasets, challenges, competitions and benchmarks. It also presents a meta-review of 12 selected deepfake-related survey papers published in 2020 and 2021, covering not only the above mentioned aspects, but also highlighting key challenges and recommendations. References [1] Darius Afchar, Vincent Nozick, Junichi Yamagishi, and Isao Echizen. 2018. MesoNet: A Compact Facial Video Forgery Detection Network. In . IEEE, 1–7. Proceedings of the 2018 IEEE International Workshop on Information Forensics and Security https://doi.org/10.1109/WIFS.2018.8630 761 [2] Henry Ajder, Giorgio Patrini, Francesco Cavalli, and Laurence Cullen. 2019. The State of Deepfakes: Landscape, Threats, and Impact. Deeptrace. , 27 pages. https://sensity.ai/reports/ [3] Zahid Akhtar and Tiago H. Falk. 2017. Audio-Visual Multimedia Quality Assessment: A Comprehensive Survey. 5 (2017), 21090–21117. IEEE Access https://doi.org/10.1109/ACCESS.2017.2750918 [4] Ahmed Ali மற்றும் Steve Renals. 2018. Word Error Rate Estimation for Speech Recognition: e-WER. . Association for Computational Linguistics, 20–24. Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics https://doi.org/10.18653/v1/P18-2004 [5] Sercan O ̈ . Arık, Jitong Chen, Kainan Peng, Wei Ping, and Yanqi Zhou. 2018. Neural Voice Cloning with a Few Samples. . Curran Associates Inc., 10040–10050. Proceedings of the 32nd International Conference on Neural Information Processing Systems https://papers.nips.cc/paper/201 8/hash/4559912e7a94a9c32b09d894f2bc3c82-Abstract.html [6] Aayush Bansal, Shugao Ma, Deva Ramanan, and Yaser Sheikh. 2018. Recycle-GAN: Unsupervised Video Retargeting. In . Springer, 17 pages. Proceedings of the 2018 European Conference on Computer Vision https://doi.org/10.1007/978-3-030-01228-1 8 [7] Yu-xuan Bao, Tian-liang Lu, and Yan-hui Du. 2020. Overview of Deepfake Video Detection Technology. 47, 9 (2020), 283–292. Computer Science https://doi.org/10.11896/jsjkx.200400130 [8] Mikol-aj Bin´kowski, Jeff Donahue, Sander Dieleman, Aidan Clark, Erich Elsen, Norman Casagrande, Luis C. Cobo, and Karen Simonyan. 2019. High Fidelity Speech Synthesis with Adversarial Networks. https://doi.org/10.48550/ARXIV.1909.11646 [9] ![](file:///C:/Users/user/AppData/Local/Temp/msohtmlclip1/01/clip_image050.gif)Madeline Brady. 2020. Deepfakes: A New Desinformation Threat? Report by the Democracy Reporting International. , 9 pages. https://democracy-reporting.org/dri publications/de epfakes-a-new-disinformation-threat/ [10] Umur Aybars Ciftci, Ilke Demir, and Lijun Yin. 2020. FakeCatcher: Detection of Synthetic Portrait Videos using Biological Signals. (2020), 17 pages. IEEE Transactions on Pattern Analysis and Machine Intelligence https://doi.org/10.1109/TPAMI.2020.3009287 [11] Hao Dang, Feng Liu, Joel Stehouwer, Xiaoming Liu, and Anil K. Jain. 2020. On the Detection of Digital Face Manipulation. In . IEEE, 10 pages. Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition https://doi.org/10.1109/CVPR42600.2020.00582 [12] Rohan Kumar Das, Tomi Kinnunen, Wen-Chin Huang, Zhen-Hua Ling, Junichi Yamagishi, Zhao Yi, Xiaohai Tian, and Tomoki Toda. 2020. Predictions of Subjective Ratings and Spoofing Assessments of Voice Conversion Challenge 2020 Submissions. In International Speech Communication Association, 99–120. Proceedings of the Joint Workshop for the Blizzard Challenge and Voice Conversion Challenge 2020 https://doi.org/10.21437/VCC BC.2020-15 [13] H´ector Delgado, Nicholas Evans, Tomi Kinnunen, Kong Aik Lee, Xuechen Liu, Andreas Nautsch, Jose Patino, Md Sahidullah, Massimiliano Todisco, Xin Wang, and Junichi Yamagishi. 2021. ASVspoof 2021: Automatic Speaker Verification Spoofing and Countermeasures Challenge Evaluation Plan. https://www.asvspoof.org/asvspoof2021/asvspoof2021 evaluation plan.pdf [14] H´ector Delgado, Nicholas Evans, Tomi Kinnunen, Kong Aik Lee, Xuechen Liu, Andreas Nautsch, Jose Patino, Md Sahidullah, Massimiliano Todisco, Xin Wang, and Junichi Yamagishi. 2021. ASVspoof 2021 Challenge - Logical Access Database. https://doi.org/10.5281/zenodo.4837263 [15] H´ector Delgado, Nicholas Evans, Tomi Kinnunen, Kong Aik Lee, Xuechen Liu, Andreas Nautsch, Jose Patino, Md Sahidullah, Massimiliano Todisco, Xin Wang, and Junichi Yamagishi. 2021. ASVspoof 2021 Challenge - Speech Deepfake Database. https://doi.org/10.5281/zenodo.4835108 [16] ![](file:///C:/Users/user/AppData/Local/Temp/msohtmlclip1/01/clip_image051.gif)Anushree Deshmukh and Sunil B. Wankhade. 2021. Deepfake Detection Approaches Using Deep Learning: A Systematic Review. . Springer, 293–302. Intelligent Computing and Networking: Proceedings of IC-ICN 2020 (Lecture Notes in Networks and Systems, Vol. https://doi.org/ 10.1007/978-981-15-7421-4 27 [17] Xinyi Ding, Zohreh Raziei, Eric C. Larson, Eli V. Olinick, Paul Krueger, மற்றும் Michael Hahsler. 2020, 1 (2020), 1–12 EURASIP Journal on Information Security https://doi.org/10.1186/s13635-020-00109-8 [18] Brian Dolhansky, Joanna Bitton, Ben Pflaum, Jikuo Lu, Russ Howes, Menglin Wang, and Cristian Canton Ferrer. 2020. The DeepFake Detection Challenge (DFDC) Dataset. https://doi.org/10.48550/ARXIV.2006.07397 [19] Brian Dolhansky, Joanna Bitton, Ben Pflaum, Jikuo Lu, Russ Howes, Menglin Wang, மற்றும் Cristian Canton Ferrer. 2020. The DeepFake Detection Challenge (DFDC) Dataset. arXiv:2006.07397. https://arxiv.org/abs/2006.07397 [20] Nick Dufour and Andrew Gully. 2019. Contributing Data to Deepfake Detection Research. Google AI Blog. https://ai.googleblog.com/2019/09/contributing-data-to-deepfake-detectio n.html [21] Ricard Durall, Margret Keuper, Franz-Josef Pfreundt, and Janis Keuper. 2019. Unmasking Deep-Fakes with Simple Features. https://doi.org/10.48550/ARXIV.1911.00686 [22] Cristian Canton Ferrer, Brian Dolhansky, Ben Pflaum, Joanna Bitton, Jacqueline Pan, and Jikuo Lu. 2020. Deepfake Detection Challenge Results: An Open Initiative to Advance AI. Meta AI Blog. https://ai.facebook.com/blog/deepfake-detection-challenge-results-an-open-initia tive-to-advance-ai/ [23] Gereon Fox, Wentao Liu, Hyeongwoo Kim, Hans-Peter Seidel, Mohamed Elgharib, and Christian Theobalt. 2021. Videoforensicshq: Detecting High-Quality Manipulated Face Videos. In . IEEE, 1–6. Proceedings of the 2021 IEEE International Conference on Multimedia and Expo https://doi.or g/10.1109/ICME51207.2021.9428101 [24] Haiying Guan, Andrew Delgado, Yooyoung Lee, Amy N. Yates, Daniel Zhou, Timothee Kheyrkhah, and Jon Fiscus. 2021. User Guide for NIST Media Forensic Challenge (MFC) Datasets. https://doi.org/10.6028/NIST.IR.8377 [25] Yinan He, Bei Gan, Siyu Chen, Yichun Zhou, Guojun Yin, Luchuan Song, Lu Sheng, Jing Shao, and Ziwei Liu. 2021. ForgeryNet: A Versatile Benchmark for Comprehensive Forgery Analysis. In . IEEE, 4360–4369. Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition https://doi.org/10.1109/CVPR46437.2021.00434 [26] Liming Jiang, Zhengkui Guo, Wayne Wu, Zhaoyang Liu, Ziwei Liu, Chen Change Loy, Shuo Yang, Yuanjun Xiong, Wei Xia, Baoying Chen, Peiyu Zhuang, Sili Li, Shen Chen, Taiping Yao, Shouhong Ding, Jilin Li, Feiyue Huang, Liujuan Cao, Rongrong Ji, Changlei Lu, and Ganchao Tan. 2021. DeeperForensics Challenge 2020 on Real-World Face Forgery Detection: Methods and Results. arXiv:2102.09471. https://arxiv.org/pdf/2102.09471.pdf [27] Liming Jiang, Ren Li, Wayne Wu, Chen Qian, மற்றும் Chen Change Loy. 2020. DeeperForensics-1.0: உண்மையான உலகம் முகம் பொய்யை கண்டுபிடிக்க ஒரு பெரிய அளவிலான தரவு சேவை. . IEEE, 2886–2895. Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition https://doi.org/ 10.1109/CVPR42600.2020.00296 [28] Nal Kalchbrenner, Erich Elsen, Karen Simonyan, Seb Noury, Norman Casagrande, Edward Lockhart, Florian Stimberg, Aaron van den Oord, Sander Dieleman, and Koray Kavukcuoglu. 2018. Efficient Neural Audio Synthesis. https://doi.org/10.48550/ARXIV.1802.08435 [29] Takuhiro Kaneko and Hirokazu Kameoka. 2017. Parallel-Data-Free Voice Conversion Using Cycle-Consistent Adversarial Networks. https://doi.org/10.48550/ARXIV.1711.11293 [30] Tero Karras, Samuli Laine, and Timo Aila. 2019. A Style-based Generator Architecture for Generative Adversarial Networks. In IEEE, 4401 – 4410 இல் Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition https://doi.org/10.1109/CVPR.2019.00453 [31] ரோஸ் கேலி. 2021. Digital Art Exhibition to Showcase Creative Potential of AI. DIGIT News. https://www.digit.fyi/digital-art-exhibition-to-showcase-creative-potential-of-ai/ [32] Ali Khodabakhsh, Raghavendra Ramachandra, Kiran Raja, Pankaj Wasnik, and Christoph Busch. 2018. Fake Face Detection Methods: Can They Be Generalized?. In . IEEE, 1–6. Proceedings of the 2018 International Conference of the Biometrics Special Interest Group https://doi.org/10.2 3919/BIOSIG.2018.8553251 [33] Hyeongwoo Kim, Mohamed Elgharib, Hans-Peter Zoll¨ofer, Michael Seidel, Thabo Beeler, Christian Richardt, and Christian Theobalt. 2019. Neural Style-Preserving Visual Dubbing. 38, 6, Article 178 (2019), 13 pages. ACM Transactions - கிராபிக்ஸ் https://doi.org/10.1145/3355089.3356500 [34] Hyeongwoo Kim, Pablo Garrido, Ayush Tewari, Weipeng Xu, Justus Thies, Matthias Niessner, Patrick P´erez, Christian Richardt, Michael Zollh¨ofer, and Christian Theobalt. 2018. Deep Video Portraits. 37, 4, Article 163 (2018), 14 pages. ACM Transactions - கிராபிக்ஸ் https://doi.org/ 10.1145/3197517.3201283 [35] Pavel Korshunov மற்றும் S'ebastien Marcel. 2019. Vulnerability Assessment and Detection of Deepfake Videos. . IEEE, 1–6. 2019 International Conference on Biometrics இன் செயல்முறைகள் https://doi.org/10.1109/ICB45273.2019.8987375 [36] Patrick Kwon, Jaeseong You, Gyuhyeon Nam, Sungwoo Park, and Gyeongsu Chae. 2021. KoDF: A Large-scale Korean DeepFake Detection Dataset. . IEEE, 10724–10733. 2021 IEEE/CVF International Conference on Computer Vision இன் திட்டங்கள் https://doi.org/10.1109/ICCV 0 0 0 0 0 0 0 0 0 0 [37] Lingzhi Li, Jianmin Bao, Hao Yang, Dong Chen, and Fang Wen. 2020. FaceShifter: Towards High Fidelity And Occlusion Aware Face Swapping. arXiv:1912.13457. https://arxiv.org/abs/1912.13457 [38] Xurong Li, Shouling Ji, Chunming Wu, Zhenguang Liu, Shuiguang Deng, Peng Cheng, Min Yang, மற்றும் Xiangwei Kong. 32, 2 (2021), 496–518. Software நாவல்கள் http://www.jos.org.cn/1000-9825/6140.htm [39] Yuezun Li, Ming-Ching Chang, and Siwei Lyu. 2018. In Ictu Oculi: Exposing AI Created Fake Videos by Detecting Eye Blinking. In ஐபிஎல், 1 - 7 IEEE International Workshop on Information Forensics and Security 2018 இல் அறிக்கைகள் https://doi.org/10.1109/WIFS.2018.8630787 [40] Yuezun Li, Xin Yang, Pu Sun, Honggang Qi, மற்றும் Siwei Lyu. 2020. Celeb-DF: DeepFake Forensics க்கான பெரிய அளவிலான பிரச்னை தரவு சேர்க்கை. . IEEE, 3204–3213. IEEE/CVF Conference on Computer Vision and Pattern Recognition (பொதுவாக) https://doi.org/10.1109/CVPR42600. 2020.00327 [41] Ruigang Liang, Peizhuo Lv, Yue Zhao, Peng Chen, Hao Xing, Yingjun Zhang, Jizhong Han, Ran He, Xianfeng Zhao, Ming Li, and Kai Chen. 2020. A Survey of Audiovisual Deepfake Detection Techniques. 5, 2 (2020), 1–17. Journal of Cybersecurity பற்றிய தகவல்கள் http://jcs.iie.ac.cn/xxaqxb/ch/re ader/view abstract.aspx?file no=20200202&flag=1 [42] Steven R. Livingstone and Frank A. Russo. 2018. The Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS): A Dynamic, Multimodal Set of Facial and Vocal Expressions in North American English. 13, 5 (2018), 35 pages. பிளஸ் 1 [43] Chen-Chou Lo, Szu-Wei Fu, Wen-Chin Huang, Xin Wang, Junichi Yamagishi, Yu Tsao, and Hsin-Min Wang. 2021. MOSNet: Deep Learning based Objective Assessment for Voice Conversion. arXiv:1904.08352. https://arxiv.org/pdf/1904.08352.pdf [44] Jaime Lorenzo-Trueba, Junichi Yamagishi, Tomoki Toda, Daisuke Saito, Fernando Villavicencio, Tomi Kinnunen, and Zhenhua Ling. 2018. The Voice Conversion Challenge 2018: Promoting Development of Parallel and Nonparallel Methods. In . International Speech Communication Association, 195–202. Proceedings of the Odyssey 2018 The Speaker and Language Recognition Workshop https://doi.org/10.21437/Odyssey.2018-28 [45] Siwei Lyu.2020 Deepfake Detection: Current Challenges and Next Steps. . IEEE, 6 pages. 2020 IEEE International Conference on Multimedia Expo அலுவலகங்கள் https://doi.org/10.1109/ICMEW46912.2020.9105991 [46] Yisroel Mirsky மற்றும் Wenke Lee. 2021. The Creation and Detection of Deepfakes: A Survey. 54, 1, Article 7 (2021), 41 பக்கங்கள். ACM Computing Survey https://doi.org/10.1145/3425780 ஆங்கிலத்தில் இதை Single Orgasm, Multiple Orgasm என்றும் கூறுகிறார்கள். 22, 8 (2015), 1006–1010. IEEE Signal Processing Letters இல் பதிவிறக்க https://doi.org/10.1109/LSP.2014.2379648 [48] Jo˜ao C. Neves, Ruben Tolosana, Ruben Vera-Rodriguez, Vasco Lopes, Hugo Proen¸ca, and Julian Fierrez. 2020. GANprintR: Improved Fakes and Evaluation of the State of the Art in Face Manipulation Detection. 14, 5 (2020), 1038–1048. IEEE Journal of Selected Topics in Signal Processing https://doi.org/10.1109/JSTSP.2020.3007250 [49] NIST Media Forensics Challenge Team. 2021. Open Media Forensics Challenge 2020-2021 மதிப்பீடு திட்டம். https://mig.nist.gov/MFC/Web/EvalPlan2020/OpenMFC2020EvaluationPlan.pdf [50] Aaron van den Oord, Sander Dieleman, Heiga Zen, Karen Simonyan, Oriol Vinyals, Alex Graves, Nal Kalchbrenner, Andrew Senior, and Koray Kavukcuoglu. 2016. WaveNet: A Generative Model for Raw Audio. https://doi.org/10.48550/ARXIV.1609.03499 [51] Debajyoti Pal and Tuul Triyason. 2018. A Survey of Standardized Approaches to the Quality of Experience Evaluation for Video Services: An ITU Perspective. 2018, கட்டுரை 1391724 (2018), 25 பக்கங்கள். International Journal of Digital Multimedia Broadcasting https://doi.org/10.1155/20 18/1391724 [52] பீன் பிங், ஹிங்ஸிங் ஃபான், ஸ்பெக்டாப், ஜிங் டாங், ஜிங் டாங், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் ஸ்பெக்டாப், ஸ்பெக்டாப் https: //arxiv.org/abs/2106.01217 [53] Ivan Perov, Daiheng Gao, Nikolay Chervoniy, Kunlin Liu, Sugasa Marangonda, Chris Um´e, Mr. Dpfks, Carl Shift Facenheim, Luis RP, Jian Jiang, Sheng Zhang, Pingyu Wu, Bo Zhou, and Weiming Zhang. 2020. DeepFaceLab: Integrated, Flexible and Extensible Face-swapping Framework. https://doi.org/10.48550/ARXIV.2005.05535 [54] Stanislav Pidhorskyi, Donald A. Adjeroh, and Gianfranco Doretto. 2020. Adversarial Latent Au-toencoders. In ஐபிஎல், 10 பதிவுகள் IEEE/CVF Conference on Computer Vision and Pattern Recognition (பொதுவாக) https://doi.org/10.1109/CVPR42600.2020.01411 [55] Andreas R ̈ossler, Davide Cozzolino, Luisa Verdoliva, Christian Riess, Justus Thies, மற்றும் Matthias Nießner. 2018. FaceForensics: A Large-scale Video Dataset for Falgery Detection in Human Faces. https://doi.org/10.48550/ARXIV.1803.09179 [56] Andreas R ̈ossler, Davide Cozzolino, Luisa Verdoliva, Christian Riess, Justus Thies, மற்றும் Matthias Nießner. 2019 FaceForensics++: அறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகுறிகள். . IEEE, 1–11. 2019 International Conference on Computer Vision (பொதுவாக கணினி பார்வையைப் பற்றி) https://doi.org/10.110 0 0 0 0 0 0 0 0 0 0 0 0 0 [57] Anindya Sen. 2021. Art and Artificial Intelligence: How ‘Deepfakes’ Can Help Create Authentic Museum Experiences! Medium Blog. https://medium.com/art-world-zen/art-and-artificia l-intelligence-how-deepfakes-can-help-create-authentic-museum-experiences-65d8aa 7da29c [58] Kou Tanaka, Hirokazu Kameoka, Takuhiro Kaneko, and Nobukatsu Hojo. 2019. WaveCycleGAN2: Time-domain Neural Post-filter for Speech Waveform Generation. https://doi.org/10.48550/A ஐரோப்பியன்1904.02892 [59] Jianhua Tao, Ruibo Fu, Jiangyan Yi, Chenglong Wang, and Tao Wang. 2020. Development and Challenge of Speech Forgery and Detection. 5, 2 (2020), 28–38. Journal of Cybersecurity பற்றிய தகவல்கள் http://jcs.iie.ac.cn/xxaqxb/ch/reader/view abstract.aspx?file no=20200204&flag=1 [60] Tencent.2020 இன் Artificial Intelligence White Paper. https://tech.sina.com.cn/roll/2020-07-14/doc-iivhvpwx5201226.shtml [61] Justus Thies, Michael Zollh'ofe, மற்றும் Matthias Niessner. 2019 விரைவில் Neural Rendering: Image Synthesis using Neural Textures. 38 , 66 (2019) , 12 பக்கங்கள். ACM Transactions on Graphics https://doi.org/10.1145/3306346.3323035 [62] Tomoki Toda, Ling-Hui Chen, Daisuke Saito, Fernando Villavicencio, Mirjam Wester, Zhizheng Wu, and Junichi Yamagishi. 2016. The Voice Conversion Challenge 2016. In International Speech Communication Association, 1632–1636 இல் Interspeech 2016 விவரங்கள் https://doi.org/10.21437/Interspeech.2016-1066 [63] Massimiliano Todisco, Xin Wang, Ville Vestman, Md Sahidullah, Hector Delgado, Andreas Nautsch, Junichi Yamagishi, Nicholas Evans, Tomi Kinnunen, and Kong Aik Lee. 2019. ASVspoof 2019: Future Horizons in Spoofed and Fake Audio Detection. arXiv:1904.05441. https://arxiv.org/ pdf/1904.05441.pdf [64] Ruben Tolosana, Ruben Vera-Rodriguez, Julian Fierrez, Aythami Morales, and Javier Ortega-Garcia. 2020. Deepfakes and beyond: A Survey of face manipulation and fake detection. 64 (2020), 131–148 Information Fusion https://doi.org/10.1016/j.inffus.2020.06.014 [65] Xin Tong, Luona Wang, Xiaoqin Pan, and Jingya Wang. 2020. An Overview of Deepfake: The Sword of Damocles in AI. In . IEEE, 265–273. 2020 International Conference on Computer Vision, Image and Deep Learning - கணினி பார்வை, புகைப்படம் மற்றும் Deep Learning https://doi.org/10.1109/CVIDL51233.2020.00-88 [66] Luisa Verdoliva.2020 Media Forensics and DeepFakes: ஒரு ஆய்வு. 14 ஐந்து (2020), 910–932 IEEE Journal of Selected Topics in Signal Processing இன் அறிமுகம் https://doi.org/10.1109/JSTSP.2020.3002101 [67] சீனா, ஜூனியி Yamagishi, மெஸ்மிலியோ Todisco, ஹெக்டர் டெல்ஜோ, Andreas Nautsch, Nicholas Evans, Md Sahidullah, Ville Vestman, Tomi Kinnunen, Kong Aik Lee, Lauri Juvela, Paavo Alku, Yu-Huai Peng, Hsin-Te Hwang, Yu Tsao, Hsin-Min Wang, S'ebastien Le Maguer, Markus Becker, Fergus Henderson, Rob Clark, Yu Zhang, Quan Wang, Ye Jia, Kai Onuma, Koji Mushika, Takashi Kaneda, Yuan Jiang, Li-Juan Liu, Yi-Chiao Wu, Wen-Chin Huang, Tomoki Toda, Kou Tanaka, Hirazu Kameoka, Ingmar Steiner, Driss Matro 64 (2020), 27 பக்கங்கள் கணினி பேச்சு & Language https://doi.org/10.1016/j.csl.2020.101114 [68] Mirjam Wester, Zhizheng Wu, மற்றும் Junichi Yamagishi. 2016 இல் Voice Conversion Challenge 2016 ஆய்வு முடிவுகள். . International Speech Communication Association, 1637–1641. Interspeech 2016 அலுவலகம் https://doi.org/10.21437/Interspeech.201 6 - 1331 [69] Junichi Yamagishi, Massimiliano Todisco, Md Sahidullah, H ́ector Delgado, Xin Wang, Nicholas Evans, Tomi Kinnunen, Kong Aik Lee, Ville Vestman, and Andreas Nautsch. 2019. ASVspoof 2019: Automatic Speaker Verification Spoofing and Countermeasures Challenge Evaluation Plan. https://www.asvspoof.org/asvspoof2019/asvspoof2019 evaluation plan.pdf [70] Zhao Yi, Wen-Chin Huang, Xiaohai Tian, Junichi Yamagishi, Rohan Kumar Das, Tomi Kinnunen, Zhen-Hua Ling, and Tomoki Toda. 2020. Voice Conversion Challenge 2020 – Intra-lingual Semi-parallel and Cross-lingual Voice Conversion –. . International Speech Communication Association, 80–98. Blizzard Challenge and Voice Conversion Challenge 2020 இல் நடந்த ஒப்பந்தம் https://doi.org/10.21437/VCC BC.2020-14 [71] Mohammed A. Younus and Taha M. Hasan. 2020. Abbreviated View of Deepfake Videos Detection Techniques. In IEEE, 115 – 120 2020 ஆம் ஆண்டில் 6வது International Engineering Conference https://doi.org/10.1109/IEC49899.2020.9122916 [72] Guangtao Zhai and Xiongkuo Min. 2020 Perceptual Image Quality Assessment: A Survey. 63, Article 211301 (2020), 52 pages. அறிவியல் China Information Science https://doi.org/10.1007/s11432-019-2757-1 [73] Teng Zhang, Lirui Deng, Liang Zhang, மற்றும் Xianglei Dang. 2020. Deep Learning in Face Synthesis: A Survey on Deepfakes. . IEEE, 67–70. Proceedings of the 2020 IEEE 3rd International Conference on Computer and Communication Engineering Technology https://doi.org/10.1109/CCET5090 1.2020.9213159 இல் [74] Yuxuan Zhang, Huan Ling, Jun Gao, Kangxue Yin, Jean-Francois Lafleche, Adela Barriuso, Antonio Torralba, மற்றும் Sanja Fidler. 2021. DatasetGAN: குறைந்த மனித முயற்சியால் எளிமையான அட்டவணை தரவு தொழிற்சாலை. IEEE, 10140–10150 அளவு 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition இன் செயல்பாடுகள் https://doi.org/10.1109/CVPR46437.2021.01001 [75] ![](file:///C:/Users/user/AppData/Local/Temp/msohtmlclip1/01/clip_image051.gif)ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான்ஜான் . Springer, 70–85. 2020 European Conference on Computer Vision (அரசியல் கண்காணிப்பு) https://doi.org/10.1007/97 8-3-030-58610-2 5 [76] Yuanhan Zhang, Zhenfei Yin, Jing Shao, Ziwei Liu, Shuo Yang, Yuanjun Xiong, Wei Xia, Yan Xu, Man Luo, Jian Liu, Jianshu Li, Zhijun Chen, Mingyu Guo, Hui Li, Junfu Liu, Pengfei Gao, Tianqi Hong, Hao Han, Shijie Liu, Xinhua Chen, Di Qiu, Cheng Zhen, Dashuang Liang, Yufeng Jin, மற்றும் Zhanlong Hao. https://arxiv.org/pdf/2102.12642.pdf [77] Bo Zhao, Shaozeng Zhang, Chunxue Xu, Yifan Sun, and Chengbin Deng. 2021. Deep Fake Geography? 48, 4 (2021), 338–352. Cartography and Geographic Information Science https://doi.org/10.1080/15230406.2021.1910075 [78] Peng Zhou, Xintong Han, Vlad I. Morariu, மற்றும் Larry S. Davis. 2017 அறுவை சிகிச்சைக்கு இரண்டு வாகனங்கள் வாகனங்கள். IEEE, 1831–1839 இல் 2017 IEEE Conference on Computer Vision and Pattern Recognition விருதுகள் https://doi.org/10.1109/CVPRW.2017.229 [79] Tianfei Zhou, Wenguan Wang, Zhiyuan Liang, மற்றும் Jianbing Shen. 2021. வாயில் முகம் Forensics. IEEE, 5774 – 5784 இல் 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition இன் செயல்பாடுகள் https://doi.org/10.1109/CVPR46437.2021.00572 [80] Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros. 2017 Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks. IEEE, 2242–2251 ஆகியவை IEEE International Conference on Computer Vision 2017 இன் அறிமுகம் https://doi.org/10.1109/ICCV.2 017.244 [81] Bojia Zi, Minghao Chang, Jingjing Chen, Xingjun Ma, மற்றும் Yu-Gang Jiang. 2020 WildDeepfake: Deepfake கண்டுபிடிப்புக்கான ஒரு கடினமான உண்மையான உலகளாவிய தரவு சேர்க்கை. . ACM, 2382–2390. 28th ACM International Conference on Multimedia 2020 இல் நடந்த ஆய்வு https://doi.org/10.1145/339417 1.3413769 இந்த ஆவணங்கள் CC by 4.0 Deed (Attribution 4.0 International) License கீழ் கிடைக்கின்றன. இந்த ஆவணங்கள் CC by 4.0 Deed (Attribution 4.0 International) License கீழ் கிடைக்கின்றன.