Mozio

Kling v2.6 Motion Control turns reference materials into directed performances—ideal for creators, cutscenes, and virtual influencers whoneed consistent motion.

Kling v2.6 Motion Control: The “Do Exactly This” Text-to-Video Model

Learn about complex-valued ODEs, Riccati equations, and non-Euclidean dynamics for advanced Geometric Deep Learning.

Generalizing Kuramoto Models: Collective Motion on Lie Groups and Spheres in Machine Learning

This is a Plain English Papers summary of a research paper called <a href="https://www.aimodels.fyi/papers/arxiv/hy-motion-10-scaling-flow-matching-models?utm_source=hackernoon&utm_medium=referral">HY-Motion 1.0: Scaling Flow Matching Models for Text-To-Motion Generation</a>. If you like these kinds of analysis, join <a href="https://www.aimodels.fyi/?utm_source=hackernoon&utm_medium=referral">AIModels.fyi</a> or follow us on <a href="https://x.com/aimodelsfyi">Twitter</a>.

<h2>The motion generation problem</h2>
For years, generating realistic human motion from text descriptions has felt stuck. Current models either fail to understand what you're asking for or produce movement that looks jerky and unnatural. Ask for an "angry walk toward a door," and the model might generate walking that's roughly the right speed but misses the emotional quality. Ask for something specific like "athletic jump with both arms extended," and it often collapses entirely. The fundamental challenge is that motion has temporal structure, physical constraints, and an almost infinite solution space. Unlike generating a static image where pixels either look right or wrong, motion requires the model to understand not just the shape of movement, but how emotion deforms it, how intention curves trajectories, and how multiple text concepts combine into a single coherent sequence.
This is why every model released so far has struggled with instruction-following. They catch maybe 70% of what you asked for and miss the nuance. The problem isn't that researchers don't understand the algorithms well enough. The bottleneck is something deeper: models trained at small scale simply don't develop the ability to understand and follow detailed instructions the way language models or image generators do.
<h2>The scaling hypothesis</h2>
The past five years of AI progress have been driven almost entirely by scaling. GPT-2 at 1.5 billion parameters could barely write coherent paragraphs. Increase that scale tenfold and something shifts. The model doesn't just do the same thing slightly better. It develops new capabilities. It reasons about edge cases it never encountered. It understands nuance and context in ways that feel qualitatively different from smaller versions.
The question for motion generation is straightforward: does this pattern hold? Or is something fundamentally different about this problem that makes scaling unhelpful?
HY-Motion answers that question by testing the hypothesis directly. Build a billion-parameter motion generation model and train it properly, and it develops instruction-following capabilities that smaller models never achieve. A small model learns to generate common motions competently. A billion-parameter model learns to listen to instructions, to combine concepts flexibly, to handle rare motion combinations and specific constraints. The research reveals that motion generation follows the same scaling laws as language and image generation, but only under one crucial condition: you need the right training data and the right training strategy.
<h2>Building the right foundation</h2>
Scaling only works if you have high-quality data to scale on. This is the unsexy part of the paper, the part many researchers skip, but it's actually where much of the breakthrough lives.
The fundamental problem is that motion datasets are messy. Raw motion capture contains jitter and artifacts from the recording process. Text descriptions are often vague or incorrect. Without cleaning, any model trained on this noise learns garbled patterns. HY-Motion treats data as a first-class problem.
<a href="https://arxiv.org/html/2512.23464/x2.png"><img src="https://arxiv.org/html/2512.23464/x2.png" alt="Data processing pipeline overview" /></a> Data processing pipeline overview
The data processing pipeline shows how raw motion capture data flows through cleaning, annotation, and quality control stages
The processing pipeline performs rigorous motion cleaning to remove artifacts and temporal inconsistencies. Careful captioning ensures text actually describes the motion rather than being generic labels. The team then organized motions into a hierarchical structure that gives the model rich conceptual structure to learn from.
<a href="https://arxiv.org/html/2512.23464/x3.png"><img src="https://arxiv.org/html/2512.23464/x3.png" alt="Motion category hierarchy spanning 200+ categories across 6 major classes" /></a> Motion category hierarchy spanning 200+ categories across 6 major classes
The hierarchy shows how motions are organized: 6 major classes branch into 200+ specific categories, giving the model granular conceptual structure
This hierarchical organization isn't arbitrary. It reflects how motion actually structures itself in human understanding. The model learns not just individual motions, but relationships between them. How does walking differ from running? How does emotion modulate both? The cleaned dataset spans over 3,000 hours of motion data, and another 400 hours gets reserved for high-quality fine-tuning. This foundation is what makes scaling meaningful. Without it, you'd train a billion-parameter model on garbage.
<h2>The three-stage training recipe</h2>
Large-scale models are brittle without the right training strategy. HY-Motion's real innovation is the systematic approach that stabilizes and aligns them. This recipe mirrors how state-of-the-art language models are trained, bringing proven techniques from another domain into motion generation.
Pretraining on large-scale unsupervised data is where the model learns the grammar of motion. Show it 3,000 hours of diverse human movement, and it learns that walking has rhythm, that reaching has velocity constraints, that expressions change how bodies move. The model isn't trying to follow instructions yet. It's learning what plausible motion looks like, building rich internal representations of movement dynamics.
High-quality fine-tuning comes next. After pretraining, take that much smaller dataset of 400 exceptionally curated motion-text pairs. The model has already learned how bodies move. Now it learns to listen to what's being asked. This stage is where instruction-following gets tight. You want perfect alignment between text and motion here, so the model learns what precision feels like. This is the domain expert phase, where a smaller dataset of perfect examples shapes the model's behavior.
Reinforcement learning from both human feedback and reward models completes the loop. Rather than assuming fine-tuning labels are perfect, the paper adds a feedback mechanism. Humans and learned reward models evaluate whether generated motions actually match text descriptions. The model then learns to optimize against this feedback. This is where the model learns to correct its own mistakes and push past previous limitations.
<a href="https://arxiv.org/html/2512.23464/x4.png"><img src="https://arxiv.org/html/2512.23464/x4.png" alt="Overview of the HY-Motion framework showing all three training stages" /></a> Overview of the HY-Motion framework showing all three training stages
The framework diagram shows how pretraining, fine-tuning, and RLHF stages feed into each other, creating a complete training pipeline
This three-stage approach is the insight that transfers beyond motion generation. Pretraining on massive unsupervised data creates general capability. Fine-tuning on curated data tightens behavior. Alignment via feedback ensures the model does what humans actually want. The same pattern has worked for language models and shows similar promise here.
<h2>The architecture that makes scaling work</h2>
All of this depends on an architecture that actually scales well. HY-Motion uses flow matching combined with a Transformer-based architecture called DiT (Diffusion Transformer).
Flow matching is a newer generative modeling approach that's simpler and more efficient than traditional diffusion. Instead of gradually adding noise and learning to reverse it, flow matching learns smooth paths from simple distributions directly to data. When combined with a Transformer architecture proven to work at billion-parameter scale in other domains, you get something that scales beautifully. Transformers handle sequential data naturally, and flow matching gives you cleaner gradients and better sample efficiency than alternatives.
<a href="https://arxiv.org/html/2512.23464/x5.png"><img src="https://arxiv.org/html/2512.23464/x5.png" alt="Model architecture of HY-Motion DiT" /></a> Model architecture of HY-Motion DiT
The architecture diagram shows how transformer layers process motion tokens while the flow matching objective guides generation
The choice matters because it determines whether scaling actually helps or just makes things slower. Flow matching is the right algorithmic fit for motion, the way transformer architectures were the right fit for language.
<h2>Breaking the ceiling</h2>
After building the data foundation, designing the training recipe, and choosing the right architecture, the question is simple: does it work?
<a href="https://arxiv.org/html/2512.23464/x1.png"><img src="https://arxiv.org/html/2512.23464/x1.png" alt="Comparison of HY-Motion 1.0 to state-of-the-art models including DART, LoM, GoToZero, and MoMask" /></a> Comparison of HY-Motion 1.0 to state-of-the-art models including DART, LoM, GoToZero, and MoMask
Top: Quantitative comparison shows HY-Motion outperforming current open-source benchmarks. Bottom: Visual examples of HY-Motion outputs demonstrate natural and precise motion generation
The difference between HY-Motion and previous models isn't incremental. It's qualitative. Previous models generate motion that vaguely resembles what you asked for, but specificity breaks them. Ask for an angry walk with both arms down, and they struggle. Ask for motion with fine-grained athletic requirements, and they often fail. HY-Motion actually listens.
<a href="https://arxiv.org/html/2512.23464/x6.png"><img src="https://arxiv.org/html/2512.23464/x6.png" alt="Visual comparison examples showing HY-Motion versus state-of-the-art alternatives" /></a> Visual comparison examples showing HY-Motion versus state-of-the-art alternatives
Side-by-side examples show how HY-Motion produces more natural, anatomically correct motion that more precisely matches text descriptions
The model has developed instruction-following capabilities that smaller models never achieved. It understands motion at the level of detail that matters for real applications. The coverage is extensive too, spanning over 200 motion categories across 6 major classes. This breadth is what gives the model flexibility to handle novel combinations of concepts.
<h2>Implications for the future</h2>
HY-Motion demonstrates something important: motion generation follows the same scaling laws as language and image generation. This matters beyond motion specifically. It shows that when you pick a new domain and want to unlock new capabilities, the path is predictable. Gather massive amounts of data. Clean it meticulously. Train at scale. Refine with human feedback.
This also signals where motion generation is heading. For years it's been a research curiosity, interesting but not yet practical. HY-Motion represents a transition point where the technology is starting to move toward commercial maturity. That's not because of a fundamental algorithmic breakthrough, but because someone built the data pipeline and training recipe that makes scaling work.
For researchers working in other domains where scaling hasn't yet unlocked capabilities, the lesson is clear: look first at your data and training strategy before designing new architectures. The bottleneck is usually not the algorithm.
The open-source release of HY-Motion is significant too. Previous motion generation work lived in closed research codebases. Making this available accelerates the entire field. Other researchers can build on this foundation, test new ideas, and push the boundary forward faster than any single lab could alone.
The broader pattern here connects to work on <a href="https://aimodels.fyi/papers/arxiv/text-driven-motion-generation-overview-challenges-directions?utm_source=hackernoon&utm_medium=referral">text-driven motion generation</a> that identified core challenges in the field, and research like <a href="https://aimodels.fyi/papers/arxiv/go-zero-towards-zero-shot-motion-generation?utm_source=hackernoon&utm_medium=referral">GoToZero</a> and <a href="https://aimodels.fyi/papers/arxiv/omnimogen-unifying-human-motion-generation-via-learning?utm_source=hackernoon&utm_medium=referral">OmnimoGen</a> that explored different architectural approaches. HY-Motion doesn't replace those efforts. It builds on them, showing that the scaling hypothesis holds when you get the fundamentals right.

<hr/>Original post: <a href="https://www.aimodels.fyi/papers/arxiv/hy-motion-10-scaling-flow-matching-models?utm_source=hackernoon&utm_medium=referral">Read on AIModels.fyi</a>

The Case for Bigger Models in Human Motion AI

Master Geometric Machine Learning by exploring Hyperbolic SVM (HSVM) optimization via SDP relaxation and the synchronization of Kuramoto phase oscillators

Advanced Machine Learning: Bridging SDP Relaxation and Collective Motion Dynamics

Why fast-moving teams lose alignment without realizing it—and how execution velocity can quietly erase shared understanding.

Motion Without Translation: Why Fast Teams Slowly Lose the Plot

MaGGIe achieves feature temporal consistency in videos using bidirectional Conv-GRU. It utilizes A100 GPUs and AdamW optimization for robust results.

Optimizing Video Matting: Curriculum Learning and Motion Blur Augmentation

This conclusion summarizes the key finding that disentangling upper and lower body motions is crucial for accurate 3D avatar generation

Disentangling Upper and Lower Body Motion: The Key Finding for Full-Body Avatar Reconstruction

This article presents an ablation study confirming that disentangling motion latents into upper and lower halves significantly enhances 3D avatar reconstruction

The Importance of Disentanglement: SAGE Outperforms Unified VQ-VAE Baselines in Full-Body Motion

This article outlines the evaluation metrics used to benchmark 3D avatar reconstruction models on the AMASS motion capture dataset.

Quantifying Motion Accuracy: MPJRE, MPJPE, and Consistency Metrics for HMD-based Avatar Generation

This article describes the online inference stage of the model, which predicts 3D avatars in a frame-by-frame manner using a sliding window approach

Temporal Refinement in Stratified Motion Diffusion: Utilizing GRU for Smoothed Full-Body Prediction

This article describes the technique for achieving disentangled motion representation by splitting full-body kinematics into upper and lower halves.

Disentangled Motion Representation: Encoding Full-Body Avatars into Discrete Latent Spaces

This article reviews the literature on full-body motion reconstruction from sparse input, tracing the evolution from IMU-based methods

Decoupling Full-Body Motion: Introducing a Stratified Approach to Solve Sparse Observation Challenge

Mozio is building the ultimate city mobility app: one app to book public transit, shuttles, limos, taxis, express trains and express buses

MOZIO

mozio.com

44-59 emps

Since 2011

MOZIO

EVERGREEN INDEX #5164

Mozio's stories on HackerNoon

Kling v2.6 Motion Control: The “Do Exactly This” Text-to-Video Model

Generalizing Kuramoto Models: Collective Motion on Lie Groups and Spheres in Machine Learning

The Case for Bigger Models in Human Motion AI

Advanced Machine Learning: Bridging SDP Relaxation and Collective Motion Dynamics

Motion Without Translation: Why Fast Teams Slowly Lose the Plot

Optimizing Video Matting: Curriculum Learning and Motion Blur Augmentation

Disentangling Upper and Lower Body Motion: The Key Finding for Full-Body Avatar Reconstruction

The Importance of Disentanglement: SAGE Outperforms Unified VQ-VAE Baselines in Full-Body Motion

Quantifying Motion Accuracy: MPJRE, MPJPE, and Consistency Metrics for HMD-based Avatar Generation

Temporal Refinement in Stratified Motion Diffusion: Utilizing GRU for Smoothed Full-Body Prediction

Disentangled Motion Representation: Encoding Full-Body Avatars into Discrete Latent Spaces

Decoupling Full-Body Motion: Introducing a Stratified Approach to Solve Sparse Observation Challenge

Mozio's Related Companies

Representing this company? Click here to claim and customize this page!

Reviews About Mozio

Mozio WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps

MOZIO

mozio.com

44-59 emps

Since 2011

MOZIO

EVERGREEN INDEX #5164

Mozio's stories on HackerNoon

Kling v2.6 Motion Control: The “Do Exactly This” Text-to-Video Model

Generalizing Kuramoto Models: Collective Motion on Lie Groups and Spheres in Machine Learning

The Case for Bigger Models in Human Motion AI

Advanced Machine Learning: Bridging SDP Relaxation and Collective Motion Dynamics

Motion Without Translation: Why Fast Teams Slowly Lose the Plot

Optimizing Video Matting: Curriculum Learning and Motion Blur Augmentation

Disentangling Upper and Lower Body Motion: The Key Finding for Full-Body Avatar Reconstruction

The Importance of Disentanglement: SAGE Outperforms Unified VQ-VAE Baselines in Full-Body Motion

Quantifying Motion Accuracy: MPJRE, MPJPE, and Consistency Metrics for HMD-based Avatar Generation

Temporal Refinement in Stratified Motion Diffusion: Utilizing GRU for Smoothed Full-Body Prediction

Disentangled Motion Representation: Encoding Full-Body Avatars into Discrete Latent Spaces

Decoupling Full-Body Motion: Introducing a Stratified Approach to Solve Sparse Observation Challenge

Mozio's Related Companies

2021

4B

2011

4B

6%

2014

2.7B

2014

500M

2015

348.4M

2017

232.6M

-1%

Representing this company? Click here to claim and customize this page!

Reviews About Mozio

Mozio WIKI

Apply to get your Tech Company News Page live on HackerNoon.com today!

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps