Ming-Yu Liu

DiffusionNFT: Online Diffusion Reinforcement with Forward Process

InfoTok: Adaptive Discrete Video Tokenizer via Information-Theoretic Compression

Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning

$PhyWorldBench$: A Comprehensive Evaluation of Physical Realism in Text-to-Video Models

Large Scale Diffusion Distillation via Score-Regularized Continuous-Time Consistency

Scenethesis: A Language and Vision Agentic Framework for 3D Scene Generation

Point-It-Out: Benchmarking Embodied Reasoning for Vision Language Models in Multi-Stage Visual Grounding

NFT: Bridging Supervised Learning and Reinforcement Learning in Math Reasoning

VoxSet: Sparse Voxel Set Tokenizer for 3D Shape Generation

PointWorld: Scaling 3D World Models for In-The-Wild Robotic Manipulation

Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator

Masked Diffusion Models are Secretly Time-Agnostic Masked Models and Exploit Inaccurate Categorical Sampling

Efficient Part-level 3D Object Generation via Dual Volume Packing

High-Quality Joint Image and Video Tokenization with Causal VAE

EdgeRunner: Auto-regressive Auto-encoder for Artistic Mesh Generation

Generalized Neighborhood Attention: Multi-dimensional Sparse Attention at the Speed of Light

One-Step Diffusion Policy: Fast Visuomotor Policies via Diffusion Distillation

One-Step Diffusion Policy: Fast Visuomotor Policies via Diffusion Distillation

Wolf: Accurate Video Captioning with a World Summarization Framework

Meshtron: High-Fidelity, Artist-Like 3D Mesh Generation at Scale