# Team VI-I2R Technical Report on EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge for Action Recognition 2021

arXiv: https://arxiv.org/abs/2206.02573 · HTML: https://mlsanity.com/paper/2206.02573/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 9 references that are in the map.
- First posted on arXiv 3 Jun 2022.

## Builds on (most cited first, 4 of 4)

- [TSM: Temporal Shift Module for Efficient Video Understanding](https://mlsanity.com/paper/1811.08383/) (2018)
- [EPIC-Fusion: Audio-Visual Temporal Binding for Egocentric Action Recognition](https://mlsanity.com/paper/1908.08498/) (2019)
- [Multi-Modal Domain Adaptation for Fine-Grained Action Recognition](https://mlsanity.com/paper/2001.09691/) (2020)
- [Temporal Attentive Alignment for Large-Scale Video Domain Adaptation](https://mlsanity.com/paper/1907.12743/) (2019)

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

- [EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge: Mixed Sequences Prediction](https://mlsanity.com/paper/2307.12837/) (2023)
- [Team VI-I2R Technical Report on EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge for Action Recognition 2022](https://mlsanity.com/paper/2301.12436/) (2023)
- [A Study on Differentiable Logic and LLMs for EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge for Action Recognition 2023](https://mlsanity.com/paper/2307.06569/) (2023)

## Abstract (arXiv)

In this report, we present the technical details of our approach to the EPIC-KITCHENS-100 Unsupervised Domain Adaptation (UDA) Challenge for Action Recognition. The EPIC-KITCHENS-100 dataset consists of daily kitchen activities focusing on the interaction between human hands and their surrounding objects. It is very challenging to accurately recognize these fine-grained activities, due to the presence of distracting objects and visually similar action classes, especially in the unlabelled target domain. Based on an existing method for video domain adaptation, i.e., TA3N, we propose to learn hand-centric features by leveraging the hand bounding box information for UDA on fine-grained action recognition. This helps reduce the distraction from background as well as facilitate the learning of domain-invariant features. To achieve high quality hand localization, we adopt an uncertainty-aware domain adaptation network, i.e., MEAA, to train a domain-adaptive hand detector, which only uses very limited hand bounding box annotations in the source domain but can generalize well to the unlabelled target domain. Our submission achieved the 1st place in terms of top-1 action recognition accuracy, using only RGB and optical flow modalities as input.