# Dual-Stage Reweighted MoE for Long-Tailed Egocentric Mistake Detection

arXiv: https://arxiv.org/abs/2509.12990 · HTML: https://mlsanity.com/paper/2509.12990/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Vision mixture-of-experts, one of 12 directions in Mixture-of-experts.
- Builds on 19 of its 23 references that are in the map.
- First posted on arXiv 16 Sep 2025.

## Placed in

- Mixture-of-experts / [Vision mixture-of-experts](https://mlsanity.com/moe/vision-mixture-of-experts/)

## Builds on (most cited first, 19 of 19)

- [The Kinetics Human Action Video Dataset](https://mlsanity.com/paper/1705.06950/) (2017)
- [ViViT: A Video Vision Transformer](https://mlsanity.com/paper/2103.15691/) (2021)
- [Sharpness-Aware Minimization for Efficiently Improving Generalization](https://mlsanity.com/paper/2010.01412/) (2020)
- [Class-Balanced Loss Based on Effective Number of Samples](https://mlsanity.com/paper/1901.05555/) (2019)
- [Long-tail learning via logit adjustment](https://mlsanity.com/paper/2007.07314/) (2020)
- [HoloAssist: an Egocentric Human Interaction Dataset for Interactive AI Assistants in the Real World](https://mlsanity.com/paper/2309.17024/) (2023)
- [BadCLIP: Dual-Embedding Guided Backdoor Attack on Multimodal Contrastive Learning](https://mlsanity.com/paper/2311.12075/) (2023)
- [ReconBoost: Boosting Can Achieve Modality Reconcilement](https://mlsanity.com/paper/2405.09321/) (2024)
- [Learning with Multiclass AUC: Theory and Algorithms](https://mlsanity.com/paper/2107.13171/) (2021)
- [Towards Robust Physical-world Backdoor Attacks on Lane Detection](https://mlsanity.com/paper/2405.05553/) (2024)
- [Multimodal Unlearnable Examples: Protecting Data against Multimodal Contrastive Learning](https://mlsanity.com/paper/2407.16307/) (2024)
- [AUCSeg: AUC-oriented Pixel-level Long-tail Semantic Segmentation](https://mlsanity.com/paper/2409.20398/) (2024)
- [DirMixE: Harnessing Test Agnostic Long-tail Recognition with Hierarchical Label Variations](https://mlsanity.com/paper/2405.07780/) (2024)
- [Size-invariance Matters: Rethinking Metrics and Losses for Imbalanced Multi-object Salient Object Detection](https://mlsanity.com/paper/2405.09782/) (2024)
- [Focal-SAM: Focal Sharpness-Aware Minimization for Long-Tailed Classification](https://mlsanity.com/paper/2505.01660/) (2025)
- [OpenworldAUC: Towards Unified Evaluation and Optimization for Open-world Prompt Tuning](https://mlsanity.com/paper/2505.05180/) (2025)
- [Not All Pairs are Equal: Hierarchical Learning for Average-Precision-Oriented Video Retrieval](https://mlsanity.com/paper/2407.15566/) (2024)
- [Improved Diversity-Promoting Collaborative Metric Learning for Recommendation](https://mlsanity.com/paper/2409.01012/) (2024)
- [Hybrid Generative Fusion for Efficient and Privacy-Preserving Face Recognition Dataset Generation](https://mlsanity.com/paper/2508.10672/) (2025)

## Built on by (most cited first, 3 of 3)

- [Guiding Diffusion-based Reconstruction with Contrastive Signals for Balanced Visual Representation](https://mlsanity.com/paper/2603.04803/) (2026)
- [Understanding-Enhanced Model Collaboration for Long-Tailed Egocentric Mistake Detection](https://mlsanity.com/paper/2606.02120/) (2026)
- [SkillSpotter: Pose-Aware Multi-View Skilled Action Detection and Grading in Ego-Exo Videos](https://mlsanity.com/paper/2606.31127/) (2026)

## Abstract (arXiv)

In this report, we address the problem of determining whether a user performs an action incorrectly from egocentric video data. To handle the challenges posed by subtle and infrequent mistakes, we propose a Dual-Stage Reweighted Mixture-of-Experts (DR-MoE) framework. In the first stage, features are extracted using a frozen ViViT model and a LoRA-tuned ViViT model, which are combined through a feature-level expert module. In the second stage, three classifiers are trained with different objectives: reweighted cross-entropy to mitigate class imbalance, AUC loss to improve ranking under skewed distributions, and label-aware loss with sharpness-aware minimization to enhance calibration and generalization. Their predictions are fused using a classification-level expert module. The proposed method achieves strong performance, particularly in identifying rare and ambiguous mistake instances. The code is available at https://github.com/boyuh/DR-MoE.