Research map / Video / Video segmentation and tracking
Facial Expression Recognition with Adaptive Frame Rate based on Multiple Testing Correction
Andrey Savchenko
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 |
|---|---|---|---|
| Benjamini-Hochberg multiple-testing thresholds instead of the same FAR at every step (ablation) | adaptive frame-rate pipeline with EmotiEffNet-B0 · ABAW-3 validation (200->100->50->10->1) | F1-score and time per frame: F1 0.4190 -> 0.4217; time 20.15 -> 3.70 ms | Improved |
| Adaptive frame-rate selection with BH-corrected thresholds vs. processing all frames | EmotiEffNet-B0 · ABAW-3 validation | F1-score: 0.4262 -> 0.4217 (drop under 0.5%) | Mixed |
| Adaptive frame-rate selection with BH thresholds vs. all frames | EmotiEffNet-B0 · AFEW validation (18->9->1) | accuracy: 0.5927 -> 0.5850 | Hurt |
| Adaptive frame-rate selection with BH thresholds vs. all frames | MT-EmotiEffNet-B0 and EmotiEffNet-B2 · ABAW-3 validation | speed-up over all frames: up to 40x (MT-EmotiEffNet) and 25x (EmotiEffNet-B2) | Improved |
| Proposed adaptive method vs. FrameExit early exit | EmotiEffNet-B0 · ABAW-3 validation | F1-score and time: F1 0.4217 vs 0.4177; time 3.70 vs 5.97 ms | Improved |
| Proposed adaptive method vs. OCSampler and AR-Net | EmotiEffNet-B0 · AFEW validation | accuracy: 0.5910 vs 0.5530 (OCSampler), 0.5526 (AR-Net) | Improved |
| Proposed adaptive method vs. AdaFrame, LiteEval on ABAW | EmotiEffNet-B0 · ABAW-3 validation | F1-score: proposed 0.4217 vs AdaFrame 0.4205 and LiteEval 0.4220 (LiteEval slightly higher) | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 on most shared results.
- OCSampler — OCSampler: Compressing Videos to One Clip With Single-Step Sampling (CVPR 2022) · better on 1 of 1 shared result
- AR-Net — AR-Net: Adaptive Frame Resolution for Efficient Action Recognition (ECCV 2020) · better on 1 of 1 shared result
- LiteEval — LiteEval: A Coarse-to-Fine Framework for Resource Efficient Video Recognition (NeurIPS 2019) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- FERV39k: A Large-Scale Multi-Scene Dataset for Facial Expression Recognition in Videos (CVPR 2022 · similarity 0.87)
- Rethinking the Learning Paradigm for Dynamic Facial Expression Recognition (CVPR 2023 · similarity 0.87)
- Intensity-Aware Loss for Dynamic Facial Expression Recognition in the Wild (AAAI 2023 · similarity 0.86)
- Lifting Scheme-Based Implicit Disentanglement of Emotion-Related Facial Dynamics in the Wild (AAAI 2025 · similarity 0.84)
- Navigating Label Ambiguity for Facial Expression Recognition in the Wild (AAAI 2025 · similarity 0.84)
- Learning Adaptive Spatial Coherent Correlations for Speech-Preserving Facial Expression Manipulation (CVPR 2024 · similarity 0.84)
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.
