用人类修正模型分歧标签，提升质量又省人力。
Human-Corrected Labels Learning: Enhancing Labels Quality via Human Correction of VLMs Discrepancies
- 只对模型不一致的样本请人纠错，减少无效劳动。
- 理论推导出风险一致估计器，融合人工与模型判断。
- 在噪声标签下仍表现稳定，适合弱监督场景。
视觉-语言模型（VLMs）虽具强大生成能力，已用于数据标注，但其生成标签存在质量低（即标签噪声）和缺乏纠错机制两大缺陷。为此，我们提出人类修正标签（HCL）新范式，通过仅对存在模型分歧的样本进行高效的人类修正，实现更高品质标注并降低人力成本。我们理论上推导出一种风险一致估计器，将人类修正标签与VLM预测结合用于分类器训练；同时提出条件概率方法，利用VLM输出与模型预测联合估计标签分布。大量实验表明，该方法在分类性能上优于基线，且对标签噪声具有鲁棒性，验证了HCL在实际弱监督场景中的有效性。代码见：https://github.com/Lilianach24/HCL.git
原文摘要 · Abstract (English)
Vision-Language Models (VLMs), with their powerful content generation capabilities, have been successfully applied to data annotation processes. However, the VLM-generated labels exhibit dual limitations: low quality (i.e., label noise) and absence of error correction mechanisms. To enhance label quality, we propose Human-Corrected Labels (HCLs), a novel setting that efficient human correction for VLM-generated noisy labels. As shown in Figure 1(b), HCL strategically deploys human correction only for instances with VLM discrepancies, achieving both higher-quality annotations and reduced labor costs. Specifically, we theoretically derive a risk-consistent estimator that incorporates both human-corrected labels and VLM predictions to train classifiers. Besides, we further propose a conditional probability method to estimate the label distribution using a combination of VLM outputs and model predictions. Extensive experiments demonstrate that our approach achieves superior classification performance and is robust to label noise, validating the effectiveness of HCL in practical weak supervision scenarios. Code https://github.com/Lilianach24/HCL.git
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。