NeurIPS 2025 accepted papers on Video
235 of the 5,812 papers accepted at NeurIPS 2025 (4.0%) are on Video, compared with 3.1% at NeurIPS 2024. They fall into 4 topics.
See NeurIPS 2025 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Video generation · 92 papers
- Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion (33 citations)
- VideoREPA: Learning Physics for Video Generation through Relational Alignment with Foundation Models (11 citations)
- Force Prompting: Video Generation Models Can Learn And Generalize Physics-based Control Signals (9 citations)
- WISA: World simulator assistant for physics-aware text-to-video generation (9 citations)
- Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance (8 citations)
- MiniMax-Remover: Taming Bad Noise Helps Video Object Removal (7 citations)
- Improving Video Generation with Human Feedback (6 citations)
- Let Them Talk: Audio-Driven Multi-Person Conversational Video Generation (5 citations)
- PhysCtrl: Generative Physics for Controllable and Physics-Grounded Video Generation (5 citations)
- Señorita-2M: A High-Quality Instruction-based Dataset for General Video Editing by Video Specialists (4 citations)
- UniRelight: Learning Joint Decomposition and Synthesis for Video Relighting (4 citations)
- Video World Models with Long-term Spatial Memory (4 citations)
and 80 more in the interactive map.
Video-language understanding · 83 papers
- Video-R1: Reinforcing Video Reasoning in MLLMs (27 citations)
- Time-R1: Post-Training Large Vision Language Model for Temporal Video Grounding (9 citations)
- StreamForest: Efficient Online Video Understanding with Persistent Event Memory (7 citations)
- Vgent: Graph-based Retrieval-Reasoning-Augmented Generation For Long Video Understanding (7 citations)
- Video-RAG: Visually-aligned Retrieval-Augmented Long Video Comprehension (7 citations)
- Logic-in-Frames: Dynamic Keyframe Search via Visual Semantic-Logical Verification for Long Video Understanding (5 citations)
- ReAgent-V: A Reward-Driven Multi-Agent Framework for Video Understanding (5 citations)
- StreamBridge: Turning Your Offline Video Large Language Model into a Proactive Streaming Assistant (5 citations)
- EgoExoBench: A Benchmark for First- and Third-person View Video Understanding in MLLMs (4 citations)
- FastVID: Dynamic Density Pruning for Fast Video Large Language Models (4 citations)
- HoliTom: Holistic Token Merging for Fast Video Large Language Models (4 citations)
- Deep Video Discovery: Agentic Search with Tool Use for Long-form Video Understanding (3 citations)
and 71 more in the interactive map.
Video segmentation and tracking · 50 papers
- 4DGT: Learning a 4D Gaussian Transformer Using Real-World Monocular Videos (8 citations)
- WorldMem: Long-term Consistent World Simulation with Memory (7 citations)
- AI-Generated Video Detection via Perceptual Straightening (3 citations)
- Learning to Factorize Spatio-Temporal Foundation Models (3 citations)
- A Unified Solution to Video Fusion: From Multi-Frame Learning to Benchmarking (2 citations)
- Asymmetric Dual-Lens Video Deblurring (2 citations)
- CamSAM2: Segment Anything Accurately in Camouflaged Videos (2 citations)
- DeltaFlow: An Efficient Multi-frame Scene Flow Estimation Method (2 citations)
- Diffusion Transformers as Open-World Spatiotemporal Foundation Models (2 citations)
- Injecting Frame-Event Complementary Fusion into Diffusion for Optical Flow in Challenging Scenes (2 citations)
- Physics-Driven Spatiotemporal Modeling for AI-Generated Video Detection (2 citations)
- REFED: A Subject Real-time Dynamic Labeled EEG-fNIRS Synchronized Recorded Emotion Dataset (2 citations)
and 38 more in the interactive map.
Action recognition · 10 papers
- Disentangled Concepts Speak Louder Than Words: Explainable Video Action Recognition (2 citations)
- HopaDIFF: Holistic-Partial Aware Fourier Conditioned Diffusion for Referring Human Action Segmentation in Multi-Person Scenarios (2 citations)
- Chirality in Action: Time-Aware Video Representation Learning by Latent Straightening
- EPFL-Smart-Kitchen: An Ego-Exo Multi-Modal Dataset for Challenging Action and Motion Understanding in Video-Language Models
- Mamba Only Glances Once (MOGO): A Lightweight Framework for Efficient Video Action Detection
- Prompt-guided Disentangled Representation for Action Recognition
- Smooth Regularization for Efficient Video Recognition
- STSBench: A Large-Scale Dataset for Modeling Neuronal Activity in the Dorsal Stream of Primate Visual Cortex
- What Do Latent Action Models Actually Learn?
- Storyboard-guided Alignment for Fine-grained Video Action Recognition