Research map / Probabilistic and scientific modeling / Dynamical systems
EmWorld: Emotion World Model with Latent State Evolution for Scenario-Incremental Dynamic Facial Expression Recognition
ke Wang, Yuanyuan Liu, Kejun Liu, Yuyang Xia, Chang Tang, Yibing Zhan, Zhe Chen
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add slow-timescale (STS) scenario evolution module (ablation) | EmWorld · AI9k | Avg accuracy: 37.19 -> 40.29 | Improved |
| add fast-timescale (FTS) expression dynamics module only (ablation) | EmWorld baseline · AI9k | Avg accuracy: 37.19 -> 37.47 | Improved |
| add FTS on top of STS (ablation) | EmWorld · AI9k | Last accuracy: 38.04 -> 39.41 | Improved |
| add STS to baseline (ablation) | EmWorld baseline · SIA10k | Avg accuracy: 39.44 -> 42.10 | Improved |
| add FTS on top of STS (ablation) | EmWorld · SIA10k | Avg accuracy: 42.10 -> 42.63 | Improved |
| use EmWorld instead of DUCT | scenario-incremental DFER · FERV39k AI9k | Avg accuracy: 36.96 -> 40.56 | Improved |
| use EmWorld instead of DUCT | scenario-incremental DFER · DFEW | Avg accuracy: 43.65 -> 53.69 | Improved |
| EmWorld lowers forgetting compared with DUCT | EmWorld · AI9k | Forgetting measure (FM): 3.60 -> 3.16 | Improved |
| set Q to 0.05 instead of 0.1 (ablation) | EmWorld · AI9k | Avg accuracy: 41.53 -> 40.06 | Hurt |
| set Q to 0.2 instead of 0.1 (ablation) | EmWorld · AI9k | Avg accuracy: 41.53 -> 39.76 | Hurt |
| vary lambda_s and lambda_f around defaults (ablation) | EmWorld · AI9k | Avg accuracy: 40.37 to 41.14 | No clear effect |
| add Kalman update after each scenario | EmWorld · AI9k | training time per scene (s): 74.06 -> 75.71 | No clear effect |
Closest papers by meaning
- Affective Processes: Stochastic Modelling of Temporal Context for Emotion and Facial Expression Recognition (CVPR 2021 · similarity 0.89)
- OUS: Bridging Scene Context and Facial Features to Overcome the Rigid Cognitive Problem (AAAI 2025 · similarity 0.88)
- Lifting Scheme-Based Implicit Disentanglement of Emotion-Related Facial Dynamics in the Wild (AAAI 2025 · similarity 0.88)
- EmoNet-Face: An Expert-Annotated Benchmark for Synthetic Emotion Recognition (NeurIPS 2025 · similarity 0.88)
- Facial-R1: Aligning Reasoning and Recognition for Facial Emotion Analysis (AAAI 2026 · similarity 0.87)
- DEEPTalk: Dynamic Emotion Embedding for Probabilistic Speech-Driven 3D Face Animation (AAAI 2025 · similarity 0.87)
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.
