Data reweighting net for web fine-grained image classification

• Published in: Multimedia tools and applications Vol. 83; no. 33; pp. 79985 - 80005
• Main Authors: Liu, Yifeng, Wu, Zhenxin, Lo, Sio-long, Chen, Zhenqiang, Ke, Gang, Yue, Chuan
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2024; Springer Nature B.V
• Subjects: Aircraft; Bird impact; Birds; Classification; Computer Communication Networks; Computer Science; Data acquisition; Data Structures and Information Theory; Datasets; Deep learning; Image acquisition; Image classification; Machine learning; Multimedia; Multimedia Information Systems; Special Purpose and Application-Based Systems; Subject specialists; Track 6: Computer Vision for Multimedia Applications; Fine-grained visual classification; Noisy labels; Web images; Data Reweighting
• DOI: 10.1007/s11042-024-18598-x
• ISSN: 1573-7721; 1380-7501

Fine-grained visual classification (FGVC) necessitates expert knowledge,which is expensive and requires a large training sample size. Consequently, using sample data acquired through the web has emerged as a novel approach for augmenting training samples. However, the web data often includes noisy samples, leading to misclassification of deep learning models. This paper presents a a meta-learning-base method called Data Reweighting Net (DR-Net). It enables the use of small, clean meta set as a guiding mechanism to accurately learn web image datasets that contain noise. More specifically, the DR-Net fully learns from small, clean meta set to discard noisy samples and obtain clean web samples through low similarity properties. DR-Net enables classification networks to adaptively learn training sets through sample weighting, mitigating the impact of noisy labels on classification learning. Our experiments on Web-bird, Web-aircraft, Web-car, CIFAR-10, and CIFAR-100 datasets demonstrate the feasibility of our proposed method.