让模型用图像区域生成可解释概念，少数据也能精准分类。
DCBM: Data-Efficient Visual Concept Bottleneck Models
- 用分割/检测模型自动提取图像区域作为概念，不依赖文本描述。
- 在小样本下仍能保持高可解释性，适合细粒度与分布外任务。
- 概念由数据自动生成，适配新领域，无需预设概念集。
概念瓶颈模型（CBMs）通过基于人类可理解的概念进行预测，提升神经网络的可解释性。然而，现有CBM通常依赖大语言模型或大规模图像语料库提取概念，难以在数据稀疏场景中应用。本文提出数据高效概念瓶颈模型（DCBM），减少概念生成所需样本量，同时保持可解释性。DCBM将概念定义为由分割或检测基础模型检测出的图像区域，使每张图像可在不同粒度上生成多个概念。该方法摆脱对文本描述和大规模预训练的依赖，适用于细粒度分类与分布外任务。使用Grad-CAM的归因分析表明，DCBM可定位测试图像中的视觉概念。通过采用数据相关的概念而非预定义概念，显著提升对新领域的适应能力。
原文摘要 · Abstract (English)
Concept Bottleneck Models (CBMs) enhance the interpretability of neural networks by basing predictions on human-understandable concepts. However, current CBMs typically rely on concept sets extracted from large language models or extensive image corpora, limiting their effectiveness in data-sparse scenarios. We propose Data-efficient CBMs (DCBMs), which reduce the need for large sample sizes during concept generation while preserving interpretability. DCBMs define concepts as image regions detected by segmentation or detection foundation models, allowing each image to generate multiple concepts across different granularities. This removes reliance on textual descriptions and large-scale pre-training, making DCBMs applicable for fine-grained classification and out-of-distribution tasks. Attribution analysis using Grad-CAM demonstrates that DCBMs deliver visual concepts that can be localized in test images. By leveraging dataset-specific concepts instead of predefined ones, DCBMs enhance adaptability to new domains.
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。