An Evaluation of Action Recognition Models on EPIC-Kitchens

Abstract

We benchmark contemporary action recognition models (TSN, TRN, and TSM) on the recently introduced EPIC-Kitchens dataset and release pretrained models on GitHub (https://github.com/epic-kitchens/action-models) for others to build upon. In contrast to popular action recognition datasets like Kinetics, Something-Something, UCF101, and HMDB51, EPIC-Kitchens is shot from an egocentric perspective and captures daily actions in-situ. In this report, we aim to understand how well these models can tackle the challenges present in this dataset, such as its long tail class distribution, unseen environment test set, and multiple tasks (verb, noun and, action classification). We discuss the models' shortcomings and avenues for future research.

arXiv comment6 pages, 3 figures, 3 tables. Models released at https://github.com/epic-kitchens/action-models

References in corpus (4)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Two-Stream Convolutional Networks for Action Recognition in Videos
• The Kinetics Human Action Video Dataset
• TSM: Temporal Shift Module for Efficient Video Understanding

Cited by in corpus (5)

• Learning to Recognize Actions on Objects in Egocentric Video with Attention Dictionaries
• Knowing What, Where and When to Look: Efficient Video Action Modeling with Attention
• Symbiotic Attention with Privileged Information for Egocentric Action Recognition
• Egocentric Action Recognition by Video Attention and Temporal Context
• A Prospective Study on Sequence-Driven Temporal Sampling and Ego-Motion Compensation for Action Recognition in the EPIC-Kitchens Dataset