具有自包含置信度的网络监督图像分类
Webly Supervised Image Classification with Self-Contained Confidence
摘要：
本文主要研究网络监督学习（WSL），其中数据集是通过从互联网上抓取样本并直接使用搜索查询作为Web标签来构建的。尽管WSL受益于快速和低成本的数据汇集，但Web标签中的噪声阻碍了图像分类模型的更好性能。为了缓解这个问题，在最近的工作中，自标记监督损失$\mathcal{L}_s$与网络监督损失$\mathcal{L}_w$一起使用。$\mathcal{L}_s$依赖于模型本身预测的伪标签。由于web标签或伪标签的正确性对于每个web样本通常是基于逐个情况的，因此期望在样本级上调整$\mathcal{L}_s$和$\mathcal{L}_w$之间的平衡。受深度神经网络（DNN）在置信度预测方面的能力的启发，我们通过适应WSL设置的模型不确定性来引入自包含置信度（SCC），并使用它来样本平衡$\mathcal{L}_s$和$\mathcal{L}_w$。因此，提出了一个简单而有效的WSL框架。研究了一系列SCC友好的正则化方法，其中提出的图增强混合是提供高质量置信度以增强我们的框架的最有效方法。在WebVision-1000和Food 101-N两个大规模WSL数据集上，该框架取得了最先进的结果。代码可在https://github.com/bigvideoresearch/SCC上获得。
英文摘要：
This paper focuses on webly supervised learning (WSL), where datasets are built by crawling samples from the Internet and directly using search queries as web labels. Although WSL benefits from fast and low-cost data collection, noises in web labels hinder better performance of the image classification model. To alleviate this problem, in recent works, self-label supervised loss $\mathcal{L}_s$ is utilized together with webly supervised loss $\mathcal{L}_w$. $\mathcal{L}_s$ relies on pseudo labels predicted by the model itself. Since the correctness of the web label or pseudo label is usually on a case-by-case basis for each web sample, it is desirable to adjust the balance between $\mathcal{L}_s$ and $\mathcal{L}_w$ on sample level. Inspired by the ability of Deep Neural Networks (DNNs) in confidence prediction, we introduce Self-Contained Confidence (SCC) by adapting model uncertainty for WSL setting, and use it to sample-wisely balance $\mathcal{L}_s$ and $\mathcal{L}_w$. Therefore, a simple yet effective WSL framework is proposed. A series of SCC-friendly regularization approaches are investigated, among which the proposed graph-enhanced mixup is the most effective method to provide high-quality confidence to enhance our framework. The proposed WSL framework has achieved the state-of-the-art results on two large-scale WSL datasets, WebVision-1000 and Food101-N. Code is available at https://github.com/bigvideoresearch/SCC.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
