A Baseline Study and Benchmark for Few-Shot Open-Set Action Recognition with Feature Residual Discrimination
This work is significant for researchers and practitioners working on real-world video analysis systems that need to handle novel, unseen actions while maintaining performance on known actions, representing an incremental advancement in the field.
This paper addresses the problem of few-shot open-set action recognition (FSOS-AR) in videos, where models need to identify known actions while rejecting unknown ones with limited examples. The authors propose a Feature-Residual Discriminator (FR-Disc) that significantly improves unknown rejection without sacrificing accuracy on known classes, establishing a new state-of-the-art for FSOS-AR.
Few-Shot Action Recognition (FS-AR) has shown promising results but is often limited by a closed-set assumption that fails in real-world open-set scenarios. While Few-Shot Open-Set (FSOS) recognition is well-established for images, its extension to spatio-temporal video data remains underexplored. To address this, we propose an architectural extension based on a Feature-Residual Discriminator (FR-Disc), adapting previous work on skeletal data to the more complex video domain. Extensive experiments on five datasets demonstrate that while common open-set techniques provide only marginal gains, our FR-Disc significantly enhances unknown rejection capabilities without compromising closed-set accuracy, setting a new state-of-the-art for FSOS-AR. The project website, code, and benchmark are available at: https://hsp-iit.github.io/fsosar/.