Research map / Label-efficient and robust learning / Recognition and long-tail
Multi-Dimensional, Nuanced and Subjective - Measuring the Perception of Facial Expressions
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use a full six-primary-expression linear model rather than only the two normative component expressions (ablation) | Compound expression prediction · Annotated face images, 10-fold cross-validation | MAE (% of [0, 1] dynamic range): 7.5% -> 6.2% | Helped |
| Use a full six-primary-expression linear model rather than only the two best-fitting primary expressions (ablation) | Compound expression prediction · Annotated face images, 10-fold cross-validation | MAE (% of [0, 1] dynamic range): 7.8% -> 6.2% | Helped |
| Represent annotations with the six primary dimensions instead of all 21 primary and compound dimensions (ablation) | Facial expression annotation · ExpW face images | compound-expression prediction MAE (% of [0, 1] dynamic range): 6.2% -> 7.5% | Mixed |
| Use six primary dimensions to predict compound expressions instead of annotating compound dimensions directly | Compound expression prediction · Annotated face images | MAE (% of [0, 1] dynamic range): 6.2% | Helped |
| Use six annotators rather than fewer annotators to fit expression-rating distributions | Crowdsourced facial-expression annotation · Primary expression ratings | cross-entropy | Helped |
| Benchmark algorithms against distributional, ambiguity-aware annotations rather than only one-hot labels | Facial expression algorithm evaluation · ExpW 1,000-image subset | algorithm ranking | Helped |
| Use DAN rather than the other evaluated algorithms | Facial expression prediction · ExpW 1,000-image subset | performance across evaluated expressions and metrics | Helped |
Closest papers by meaning
- EmotioNet: An Accurate, Real-Time Algorithm for the Automatic Annotation of a Million Facial Expressions in the Wild (CVPR 2016 · similarity 0.90)
- EmoNet-Face: An Expert-Annotated Benchmark for Synthetic Emotion Recognition (NeurIPS 2025 · similarity 0.87)
- Navigating Label Ambiguity for Facial Expression Recognition in the Wild (AAAI 2025 · similarity 0.86)
- OUS: Bridging Scene Context and Facial Features to Overcome the Rigid Cognitive Problem (AAAI 2025 · similarity 0.86)
- Customizing Visual Emotion Evaluation for MLLMs: An Open-vocabulary, Multifaceted, and Scalable Approach (ICLR 2026 · similarity 0.86)
- Facial-R1: Aligning Reasoning and Recognition for Facial Emotion Analysis (AAAI 2026 · similarity 0.85)
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.