EPIC-KITCHENS-100 Unsupervised Domain Adaptation Challenge: Mixed Sequences Prediction

TL;DR

This paper introduces a novel approach for unsupervised domain adaptation in action recognition, leveraging sequence mixing, pseudo-labeling, language models, and co-occurrence filtering to improve transfer performance on the EPIC-Kitchens-100 dataset.

Contribution

It proposes a new sequence mixing and filtering method for unsupervised domain adaptation in action recognition tasks.

Findings

Achieved 2nd place in verb recognition on the leaderboard.

Achieved 4th place in noun and action recognition.

Demonstrated improved transfer results using sequence and language filtering techniques.

Abstract

This report presents the technical details of our approach for the EPIC-Kitchens-100 Unsupervised Domain Adaptation (UDA) Challenge in Action Recognition. Our approach is based on the idea that the order in which actions are performed is similar between the source and target domains. Based on this, we generate a modified sequence by randomly combining actions from the source and target domains. As only unlabelled target data are available under the UDA setting, we use a standard pseudo-labeling strategy for extracting action labels for the target. We then ask the network to predict the resulting action sequence. This allows to integrate information from both domains during training and to achieve better transfer results on target. Additionally, to better incorporate sequence information, we use a language model to filter unlikely sequences. Lastly, we employed a co-occurrence matrix to eliminate unseen combinations of verbs and nouns. Our submission, labeled as 'sshayan', can be found on the leaderboard, where it currently holds the 2nd position for 'verb' and the 4th position for both 'noun' and 'action'.