mlsanityData to 14 Oct 2026

Predicting Human Disagreement for Calibrated Dynamic Facial Expression Recognition

arXiv 2609.17130 ↗First posted 15 Sep 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Dynamic facial expression recognition (DFER) benchmarks such as DFEW provide multiple annotator votes per clip, yet most models collapse them to a majority label and cannot represent human disagreement at inference time. We propose a disagreement-aware DFER framework that trains directly on the raw annotator count vector using a Dirichlet-Multinomial likelihood. Unlike mean-only soft-label objectives, the proposed likelihood provides scale-sensitive supervision for the Dirichlet concentration while preserving the predictive mean. A separate ambiguity head predicts annotation entropy for unseen clips, and a monotone Chow-style reject rule combines predicted ambiguity, vacuity, temporal instability, and input quality for selective prediction. On DFEW, the method preserves recognition accuracy while reducing ECE by 30% and AURC by 15%, and predicted ambiguity reaches a Spearman correlation of 0.52 with the annotation entropy of test clips. The calibration and selective-prediction gains transfer to FERV39k and remain under identity- and movie-disjoint DFEW splits.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 27 of its 33 references that are in the map. First posted on arXiv 15 Sep 2026.

Builds on

27 of 33 references.

References · 27

  1. 2017On Calibration of Modern Neural Networkscited by 4,960
  2. 2022VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Trainingcited by 1,322
  3. 2018Evidential Deep Learning to Quantify Classification Uncertaintycited by 732
  4. 2018Predictive Uncertainty Estimation via Prior Networkscited by 557
  5. 2019Human uncertainty makes classification more robustcited by 235
  6. 2022The 'Problem' of Human Label Variation: On Ground Truth in Data, Modeling and Evaluationcited by 139
  7. 2020DFEW: A Large-Scale Database for Recognizing Dynamic Facial Expressions in the Wildcited by 128
  8. 2020Improving model calibration with accuracy versus uncertainty optimizationcited by 123
  9. 2018Bias-Reduced Uncertainty Estimation for Deep Neural Classifierscited by 113
  10. 2021Evidential Deep Learning for Open Set Action Recognitioncited by 96
  11. 2021Dive into Ambiguity: Latent Distribution Mining and Pairwise Uncertainty Estimation for Facial Expression Recognitioncited by 89
  12. 2021Prior and Posterior Networks: A Survey on Evidential Deep Learning Methods For Uncertainty Estimationcited by 71
15 more
  1. 2022FERV39k: A Large-Scale Multi-Scene Dataset for Facial Expression Recognition in Videoscited by 69
  2. 2022MAFW: A Large-scale, Multi-modal, Compound Affective Database for Dynamic Facial Expression Recognition in the Wildcited by 65
  3. 2023Prompting Visual-Language Models for Dynamic Facial Expression Recognitioncited by 44
  4. 2023MAE-DFER: Efficient Masked Autoencoder for Self-supervised Dynamic Facial Expression Recognitioncited by 43
  5. 2023From Static to Dynamic: Adapting Landmark-Aware Image Models for Facial Expression Recognition in Videoscited by 39
  6. 2022Uncertainty-aware Label Distribution Learning for Facial Expression Recognitioncited by 30
  7. 2024FineCLIPER: Multi-modal Fine-grained CLIP for Dynamic Facial Expression Recognition with AdaptERscited by 30
  8. 2024A Comprehensive Survey on Evidential Deep Learning and Its Applicationscited by 25
  9. 2024MMA-DFER: MultiModal Adaptation of unimodal models for Dynamic Facial Expression Recognition in-the-wildcited by 14
  10. 2024Crowd-Calibrator: Can Annotator Disagreement Inform Calibration in Subjective Tasks?cited by 13
  11. 2024Static for Dynamic: Towards a Deeper Understanding of Dynamic Facial Expressions Using Static Expression Datacited by 9
  12. 2025Robust Dynamic Facial Expression Recognitioncited by 4
  13. 2025MIDAS: Mixing Ambiguous Data with Soft Labels for Dynamic Facial Expression Recognitioncited by 4
  14. 2026Interpretable Uncertainty Routing Separating Emotion Ambiguity from Distribution Shift in Facial Expression Recognitioncited by 2
  15. 2026Confidence Calibration under Ambiguous Ground Truthcited by 1

Built on by

Not cited yet.