Text-Enhanced Zero-Shot Action Recognition: A training-free approach
This work addresses the problem of resource-intensive and biased training in zero-shot action recognition for video analysis, offering a simpler alternative, though it appears incremental as it builds on existing vision-language models.
The paper tackles the challenge of zero-shot video action recognition by proposing a training-free approach called TEAR, which uses action descriptors and contextual information to enhance performance without requiring training data or extensive resources, achieving competitive results on datasets like UCF101, HMDB51, and Kinetics-600.
Vision-language models (VLMs) have demonstrated remarkable performance across various visual tasks, leveraging joint learning of visual and textual representations. While these models excel in zero-shot image tasks, their application to zero-shot video action recognition (ZSVAR) remains challenging due to the dynamic and temporal nature of actions. Existing methods for ZS-VAR typically require extensive training on specific datasets, which can be resource-intensive and may introduce domain biases. In this work, we propose Text-Enhanced Action Recognition (TEAR), a simple approach to ZS-VAR that is training-free and does not require the availability of training data or extensive computational resources. Drawing inspiration from recent findings in vision and language literature, we utilize action descriptors for decomposition and contextual information to enhance zero-shot action recognition. Through experiments on UCF101, HMDB51, and Kinetics-600 datasets, we showcase the effectiveness and applicability of our proposed approach in addressing the challenges of ZS-VAR.