Beyond Scene Shortcuts: Action-Centric Video-Text Alignment for Fine-Grained Action Understanding
cs.CV
Submitted: 2026-11-04
Updated: 2026-11-04
Terminology
Sources
- PAINT: Paying Attention to INformed Tokens to Mitigate Hallucination in Large Vision-Language Model
- MotionLLM: Understanding Human Behaviors from Human Motions and Videos
- HALC: Object Hallucination Reduction via Adaptive Focal-Contrast Decoding
- MASH-VLM: Mitigating Action-Scene Hallucination in Video-LLMs through Disentangled Spatial-Temporal Representations
- Qwen Look Again: Guiding Vision-Language Reasoning Models to Re-attention Visual Information
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- Oops! Predicting Unintentional Action in Video
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- MotionBench: Benchmarking and Improving Fine-grained Video Motion Understanding for Vision Language Models
- FinePseudo: Improving Pseudo-Labelling through Temporal-Alignablity for Semi-Supervised Fine-Grained Action Recognition
- Enhancing Video Transformers for Action Understanding with VLM-aided Training
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild
- FAVOR-Bench: A Comprehensive Benchmark for Fine-Grained Video Motion Understanding
- HieroAction: Hierarchically Guided VLM for Fine-Grained Action Analysis
- IKOD: Image-Attention-Guided Key-Value Merging against Object Hallucination in the Decoding of Large Vision-Language Models
- Analyzing and Mitigating Object Hallucination in Large Vision-Language Models
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs