让视频模型自动拆分细粒度类别，无需重新训练。
Let's Split Up: Zero-Shot Classifier Edits for Fine-Grained Video Understanding
- 利用模型隐含结构实现零样本分类器拆分
- 在新细分类别上准确率显著提升，原有性能不变
- 适合需要快速更新标签体系的视频理解任务
视频识别模型通常基于固定分类体系，类别过粗，导致物体、方式或结果等细微差异被合并为单一标签。随着任务定义演变，这类模型难以适应新细分需求，而重新标注和训练成本高昂。为此，我们提出类别拆分任务：在不损失原有性能的前提下，将粗粒度类别精细拆分为子类别。本文提出一种零样本编辑方法，利用视频分类器的潜在组合结构，在无需额外数据的情况下揭示细粒度差异。同时发现，低样本微调虽简单但效果显著，且得益于本方法的零样本初始化。在新构建的视频类别拆分基准上的实验表明，该方法显著优于视觉-语言基线，在新拆分类别上提升准确率，同时保持原有类别性能不变。
原文摘要 · Abstract (English)
Video recognition models are typically trained on fixed taxonomies which are often too coarse, collapsing distinctions in object, manner or outcome under a single label. As tasks and definitions evolve, such models cannot accommodate emerging distinctions and collecting new annotations and retraining to accommodate such changes is costly. To address these challenges, we introduce category splitting, a new task where an existing classifier is edited to refine a coarse category into finer subcategories, while preserving accuracy elsewhere. We propose a zero-shot editing method that leverages the latent compositional structure of video classifiers to expose fine-grained distinctions without additional data. We further show that low-shot fine-tuning, while simple, is highly effective and benefits from our zero-shot initialization. Experiments on our new video benchmarks for category splitting demonstrate that our method substantially outperforms vision-language baselines, improving accuracy on the newly split categories without sacrificing performance on the rest. Project page: https://kaitingliu.github.io/Category-Splitting/.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。