---
title: "TEST-V: TEst-time Support-set Tuning for Zero-shot Video Classification"
authors:
  - "Rui Yan"
  - "Jin Wang"
  - "Hongyu Qu"
  - "Xiaoyu Du"
  - "Dong Zhang"
  - "Jinhui Tang"
  - "Tieniu Tan"
doi: "10.24963/ijcai.2025/239"
url: "https://ijcai.org/proceedings/2025/239"
pdf: "https://ijcai.org/proceedings/2025/0239.pdf"
venue: "Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"
year: 2025
pages: "2143-2151"
track: "Main Track"
keywords:
  - "Computer Vision: CV: Video analysis and understanding"
  - "Computer Vision: CV: Vision, language and reasoning"
---

# TEST-V: TEst-time Support-set Tuning for Zero-shot Video Classification

Rui Yan, Jin Wang, Hongyu Qu, Xiaoyu Du, Dong Zhang, Jinhui Tang, Tieniu Tan

*Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence*, pages 2143-2151. DOI: [10.24963/ijcai.2025/239](https://doi.org/10.24963/ijcai.2025/239)

## Abstract

Recently, adapting Vision Language Models (VLMs) to zero-shot visual classification by tuning class embedding with a few prompts (Test-time Prompt Tuning, TPT) or replacing class names with generated visual samples (support-set) has shown promising results. However, TPT cannot avoid the semantic gap between modalities while the support-set cannot be tuned. To this end, we draw on each other's strengths and propose a novel framework, namely TEst-time Support-set Tuning for zero-shot Video Classification (TEST-V). It first dilates the support-set with multiple prompts (Multi-prompting Support-set Dilation, MSD) and then erodes the support-set via learnable weights to mine key cues dynamically (Temporal-aware Support-set Erosion, TSE). Specifically, i) MSD expands the support samples for each class based on multiple prompts inquired from LLMs to enrich the diversity of the support-set. ii) TSE tunes the support-set with factorized learnable weights according to the temporal prediction consistency in a self-supervised manner to dig pivotal supporting cues for each class. TEST-V achieves state-of-the-art results across four benchmarks and shows good interpretability.