ICML 2025 accepted papers on Video
48 of the 3,342 papers accepted at ICML 2025 (1.4%) are on Video, compared with 0.9% at ICML 2024. They fall into 4 topics.
See ICML 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 · 23 papers
- How Far Is Video Generation from World Model: A Physical Law Perspective (5 citations)
- Long-Term TalkingFace Generation via Motion-Prior Conditional Diffusion Model (3 citations)
- MimicMotion: High-Quality Human Motion Video Generation with Confidence-aware Pose Guidance (2 citations)
- WorldSimBench: Towards Video Generation Models as World Simulators (2 citations)
- RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers (1 citation)
- Towards World Simulator: Crafting Physical Commonsense-Based Benchmark for Video Generation (1 citation)
- AsymRnR: Video Diffusion Transformers Acceleration with Asymmetric Reduction and Restoration
- Broadband Ground Motion Synthesis by Diffusion Model with Minimal Condition
- Ca2-VDM: Efficient Autoregressive Video Diffusion Model with Causal Generation and Cache Sharing
- Cavia: Camera-controllable Multi-view Video Diffusion with View-Integrated Attention
- Empowering World Models with Reflection for Embodied Video Prediction
- Fast Video Generation with Sliding Tile Attention
and 11 more in the interactive map.
Video-language understanding · 13 papers
- $\infty$-Video: A Training-Free Approach to Long Video Understanding via Continuous-Time Memory Consolidation
- $\mathcal{V}ista\mathcal{DPO}$: Video Hierarchical Spatial-Temporal Direct Preference Optimization for Large Video Models
- LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding
- Scaling Video-Language Models to 10K Frames via Hierarchical Differential Distillation
- Unifying Specialized Visual Encoders for Video Language Models
- VideoRoPE: What Makes for Good Video Rotary Position Embedding?
- DynaMind: Reasoning over Abstract Video Dynamics for Embodied Decision-Making
- Ex-VAD: Explainable Fine-grained Video Anomaly Detection Based on Visual-Language Models
- Fine-Grained Captioning of Long Videos through Scene Graph Consolidation
- Impossible Videos
- Improving LLM Video Understanding with 16 Frames Per Second
- Modularized Self-Reflected Video Reasoner for Multimodal LLM with Application to Video Question Answering
and 1 more in the interactive map.
Video segmentation and tracking · 8 papers
- An Analysis of Quantile Temporal-Difference Learning (3 citations)
- CSTrack: Enhancing RGB-X Tracking via Compact Spatiotemporal Features
- Learning Event Completeness for Weakly Supervised Video Anomaly Detection
- MoMa: Modulating Mamba for Adapting Image Foundation Models to Video Recognition
- PlaySlot: Learning Inverse Latent Dynamics for Controllable Object-Centric Video Prediction and Planning
- Video Prediction Policy: A Generalist Robot Policy with Predictive Visual Representations
- SpikeVideoFormer: An Efficient Spike-Driven Video Transformer with Hamming Attention and $\mathcal{O}(T)$ Complexity
- Unifying Knowledge from Diverse Datasets to Enhance Spatial-Temporal Modeling: A Granularity-Adaptive Geographical Embedding Approach