mlsanityData to 7 Oct 2026

A Baseline Study and Benchmark for Few-Shot Open-Set Action Recognition with Feature Residual Discrimination

arXiv 2603.04125 ↗First posted 4 Mar 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

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/.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 31 of its 48 references that are in the map. First posted on arXiv 4 Mar 2026.

Builds on

31 of 48 references.

References · 31

  1. 2017The Kinetics Human Action Video Datasetcited by 1,894
  2. 2022VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Trainingcited by 1,305
  3. 2017The "something something" video database for learning and evaluating visual common sensecited by 1,092
  4. 2016NTU RGB+D: A Large Scale Dataset for 3D Human Activity Analysiscited by 683
  5. 2021Open-Set Recognition: a Good Closed-Set Classifier is All You Need?cited by 351
  6. 2018Reducing Network Agnostophobiacited by 165
  7. 2017Generative OpenMax for Multi-Class Open Set Classificationcited by 157
  8. 2021Adversarial Reciprocal Points Learning for Open Set Recognitioncited by 153
  9. 2019Few-Shot Video Classification via Temporal Alignmentcited by 124
  10. 2021Evidential Deep Learning for Open Set Action Recognitioncited by 96
  11. 2021Temporal-Relational CrossTransformers for Few-Shot Action Recognitioncited by 83
  12. 2021Spatio-temporal Relation Modeling for Few-shot Action Recognitioncited by 52
19 more
  1. 2020Few-Shot Open-Set Recognition using Meta-Learningcited by 50
  2. 2023CLIP-guided Prototype Modulating for Few-shot Action Recognitioncited by 41
  3. 2023MoLo: Motion-augmented Long-short Contrastive Learning for Few-shot Action Recognitioncited by 36
  4. 2021Few-shot Open-set Recognition by Transformation Consistencycited by 29
  5. 2023Boosting Few-shot Action Recognition with Graph-guided Hybrid Matchingcited by 24
  6. 2023Navigating Open Set Scenarios for Skeleton-based Action Recognitioncited by 17
  7. 2020Task-Adaptive Negative Envision for Few-Shot Open-Set Recognitioncited by 14
  8. 2024Open-Set Recognition in the Age of Vision-Language Modelscited by 13
  9. 2023HyRSM++: Hybrid Relation Guided Temporal Set Matching for Few-shot Action Recognitioncited by 12
  10. 2023Glocal Energy-based Learning for Few-Shot Open-Set Recognitioncited by 10
  11. 2022Towards Generalized Few-Shot Open-Set Object Detectioncited by 8
  12. 2023Semantic-aware Video Representation for Few-shot Action Recognitioncited by 7
  13. 2022Few-shot Open-set Recognition Using Background as Unknownscited by 6
  14. 2022Knowledge Prompting for Few-shot Action Recognitioncited by 6
  15. 2024Dual-Adapter: Training-free Dual Adaptation for Few-shot Out-of-Distribution Detectioncited by 4
  16. 2022One-Shot Open-Set Skeleton-Based Action Recognitioncited by 3
  17. 2023Enlarging Instance-specific and Class-specific Information for Open-set Action Recognitioncited by 3
  18. 2024Large-Scale Evaluation of Open-Set Image Classification Techniquescited by 2
  19. 2024EOL: Transductive Few-Shot Open-Set Recognition by Enhancing Outlier Logitscited by 2

Built on by

Not cited yet.