TPAMI 2026 papers on Video
67 of the 880 papers published in TPAMI 2026 (7.6%) are on Video, compared with 7.9% at TPAMI 2025. They fall into 4 topics.
See TPAMI 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Video segmentation and tracking · 25 papers
- NVC-1B: Scaling up Neural Video Coding Models (9 citations)
- Dynamic in Static: Hybrid Visual Correspondence for Self-Supervised Video Object Segmentation (6 citations)
- Scanpath Prediction in Panoramic Videos via Expected Code Length Minimization (6 citations)
- UniFES: A Unified Recurrent Network for Quality Enhancement and Stabilization in Face Videos (6 citations)
- The Sound of Water: Inferring Physical Properties From Pouring Liquids (5 citations)
- Match Stereo Videos via Bidirectional Alignment (4 citations)
- EVDI++: Event-Based Video Deblurring and Interpolation via Self-Supervised Learning (3 citations)
- Causal Prompts for Open-Vocabulary Video Instance Segmentation (2 citations)
- Collaborative Feedback Discriminative Propagation for Video Super-Resolution (2 citations)
- LaverNet: Lightweight All-in-one Video Restoration via Selective Propagation (2 citations)
- Learning When and How to Update Memory for Video Object Segmentation (2 citations)
- Mask-Guided Self-Supervised Video Object Segmentation (1 citation)
and 13 more in the interactive map.
Video generation · 20 papers
- Helios: Real Real-Time Long Video Generation Model (77 citations)
- EDTalk++: Full Disentanglement for Controllable Talking Head Synthesis (6 citations)
- Think Before You Move: Latent Motion Reasoning for Text-to-Motion Generation (6 citations)
- PRIMEdit: Probability Redistribution for Instance-Aware Multi-Object Video Editing With a New Benchmark Dataset (4 citations)
- Consistent and Controllable Image Animation With Linear Motion Diffusion Transformers (3 citations)
- Harnessing Meta-Learning for Controllable Full-Frame Video Stabilization (3 citations)
- S$^{2}$VG: 3D Stereoscopic and Spatial Video Generation via Denoising Frame Matrix (3 citations)
- AnimateAnyMesh++: A Flexible Feed-Forward Framework for High-Fidelity Text-Driven Mesh Animation (2 citations)
- Learning Spatial-Temporal Coherent Correlations for Speech-Preserving Facial Expression Manipulation (2 citations)
- Action-Controlled Scale-Wise Flow Matching for Embodied World Models
- CLIP-Actor-X: Text-Driven 4D Human Avatar Generation via Cross-Modal Synthesis-Through-Optimization
- DrawMotion: Generating 3D Human Motions by Freehand Drawing
and 8 more in the interactive map.
Video-language understanding · 15 papers
- Sa2VA: Marrying SAM2 With MLLM for Dense Grounded Understanding of Images and Videos (206 citations)
- Ego-R1: Agentic Chain-of-Tool-Thought for Ultra-Long Egocentric Video Reasoning (65 citations)
- Decoupled Seg Tokens Make Stronger Reasoning Video Segmenter and Grounder (5 citations)
- Knowledge Diffusion-Based Adaptive Alignment With Hierarchical Context for Video Temporal Grounding (2 citations)
- Momentor++: Advancing Video Large Language Models With Fine-Grained Long Video Reasoning (1 citation)
- SELongVLM: Empowering Long Video Language Models With Self-Corrective Clip Selection (1 citation)
- Accelerating SAM2 With Efficient Memory Attention Module via Spatiotemporal Token Pruning
- ADVersa: Abductive Driving Accident Video Understanding
- DAAN: Mitigating Information Asymmetries for Point-Supervised Video Moment Retrieval
- Exploring Multimodal Adapters with Multi-Task Decoders for Video Action Recognition
- IntentQA: Intent Question Answering in Videos by Cognitive Context Reasoning
- Kalmanized Recursive Least Squares Attention for Streaming Video Understanding
and 3 more in the interactive map.
Action recognition · 7 papers
- Spatio-Temporal Decoupled Knowledge Compensator for Few-Shot Action Recognition (30 citations)
- XOV-Action: Toward Generalizable Open-Vocabulary Action Recognition (15 citations)
- Few-Shot Action Recognition via Intra- and Inter-Video Information Maximization (14 citations)
- MORIC: CSI Delay-Doppler Decomposition for Robust Wi-Fi-based Human Activity Recognition (5 citations)
- E2E-LOAD++: End-to-end Long-form Online Action Detection and Beyond (1 citation)
- Task Graph Maximum Likelihood Estimation for Procedural Activity Understanding in Egocentric Videos (1 citation)
- Toward Stronger 3D Human Action Representation Learning Via Efficient Spatiotemporal Decoupling
