Research map / Video / Video segmentation and tracking
A unique M-pattern for micro-expression spotting in long videos
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes a unique M-pattern movement signature to spot micro-expressions in long videos, addressing head-movement interference and fixed-reference-frame limitations of existing methods.
Problem
Micro-expression spotting in long videos is hard because ME motion is small and easily confused with global head movements; existing fixed-reference-frame methods degrade optical flow accuracy and depend heavily on facial alignment.
Approach
Exploits the unique movement pattern (M-pattern) of micro-expressions to distinguish them from other head or body movements, moving away from fixed-reference-frame optical flow and reducing reliance on facial alignment.
Experimental setup
Baselines: fixed reference frame MES methods
Closest papers by meaning
- MemFlow: Optical Flow Estimation and Prediction with Memory (CVPR 2024 · similarity 0.81)
- Occlusion-Insensitive Talking Head Video Generation via Facelet Compensation (AAAI 2025 · similarity 0.81)
- DPFlow: Adaptive Optical Flow Estimation with a Dual-Pyramid Framework (CVPR 2025 · similarity 0.81)
- TEASER: Token Enhanced Spatial Modeling for Expressions Reconstruction (ICLR 2025 · similarity 0.80)
- Synergizing Motion and Appearance: Multi-Scale Compensatory Codebooks for Talking Head Video Generation (CVPR 2025 · similarity 0.80)
- X-NeMo: Expressive Neural Motion Reenactment via Disentangled Latent Attention (ICLR 2025 · similarity 0.80)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.