ICLR 2026 accepted papers on Video
247 of the 5,358 papers accepted at ICLR 2026 (4.6%) are on Video, compared with 4.0% at ICLR 2025. They fall into 4 topics.
See ICLR 2026 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 · 135 papers
- Animating the Uncaptured: Humanoid Mesh Animation with Video Diffusion Models (1 citation)
- Vid2World: Crafting Video Diffusion Models to Interactive World Models (1 citation)
- AC-Foley: Reference-Audio-Guided Video-to-Audio Synthesis with Acoustic Transfer
- Any-to-Bokeh: Arbitrary-Subject Video Refocusing with Video Diffusion Model
- Arbitrary Generative Video Interpolation
- Astra: General Interactive World Model with Autoregressive Denoising
- AudioX: A Unified Framework for Anything-to-Audio Generation
- AUHead: Realistic Emotional Talking Head Generation via Action Units Control
- Beyond Skeletons: Learning Animation Directly from Driving Videos with Same2X Training Strategy
- BindWeave: Subject-Consistent Video Generation via Cross-Modal Integration
- BLADE: Block-Sparse Attention Meets Step Distillation for Efficient Video Generation
- BWCache: Accelerating Video Diffusion Transformers through Block-Wise Caching
and 123 more in the interactive map.
Video-language understanding · 66 papers
- The Human Brain as a Dynamic Mixture of Expert Models in Video Understanding (2 citations)
- CoT-RVS: Zero-Shot Chain-of-Thought Reasoning Segmentation for Videos (1 citation)
- Go Beyond Earth: Understanding Human Actions and Scenes in Microgravity Environments (1 citation)
- HiVid: LLM-Guided Video Saliency For Content-Aware VOD And Live Streaming (1 citation)
- StreamingVLM: Real-Time Understanding for Infinite Video Streams (1 citation)
- WorldSense: Evaluating Real-world Omnimodal Understanding for Multimodal LLMs (1 citation)
- A.I.R.: Enabling Adaptive, Iterative, and Reasoning-based Frame Selection For Video Question Answering
- CaReBench: A Fine-grained Benchmark for Video Captioning and Retrieval
- Decomposed Attention Fusion in MLLMs for Training-free Video Reasoning Segmentation
- Dynamic Reflections: Probing Video Representations with Text Alignment
- Enhancing Visual Token Representations for Video Large Language Models via Training-free Spatial-Temporal Pooling and Gridding
- EventFlash: Towards Efficient MLLMs for Event-Based Vision
and 54 more in the interactive map.
Video segmentation and tracking · 34 papers
- Advancing Complex Video Object Segmentation via Progressive Concept Construction
- Deforming Videos to Masks: Flow Matching for Referring Video Segmentation
- Efficient-SAM2: Accelerating SAM2 with Object-Aware Visual Encoding and Memory Retrieval
- Exploring Real-Time Super-Resolution: Benchmarking and Fine-Tuning for Streaming Content
- Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling
- LiFR-Seg: Anytime High-Frame-Rate Segmentation via Event-Guided Propagation
- Midway Network: Learning Representations for Recognition and Motion from Latent Dynamics
- NeMo-map: Neural Implicit Flow Fields for Spatio-Temporal Motion Mapping
- PredNext: Explicit Cross-View Temporal Prediction for Unsupervised Learning in Spiking Neural Networks
- ProstaTD: Bridging Surgical Triplet from Classification to Fully Supervised Detection
- RelayFormer: A Unified Local-Global Attention Framework for Scalable Image and Video Manipulation Localization
- Steering and Rectifying Latent representation manifolds in Frozen Multi-modal LLMs for Video Anomaly Detection
and 22 more in the interactive map.
Action recognition · 12 papers
- Action-Guided Attention for Video Action Anticipation
- EAST: Early Action Prediction Sampling Strategy with Token Masking
- EgoBrain: Synergizing Minds and Eyes For Human Action Understanding
- Invert4TVG: A Temporal Video Grounding Framework with Inversion Tasks Preserving Action Understanding Ability
- Privacy Beyond Pixels: Latent Anonymization for Privacy-Preserving Video Understanding
- Procedural Mistake Detection via Action Effect Modeling
- ViPRA: Video Prediction for Robot Actions
- Curvature-Guided Task Synergy for Skeleton based Temporal Action Segmentation
- From Pixels to Semantics: Unified Facial Action Representation Learning for Micro-Expression Analysis
- GaitSnippet: Gait Recognition Beyond Unordered Sets and Ordered Sequences
- Genie Envisioner: A Unified World Foundation Platform for Robotic Manipulation
- VideoAgentTrek: Computer-Use Pretraining from Unlabeled Videos