Wise optimisation: deep image embedding by informative pair weighting and ranked list learning. Issue 15 (18th February 2021)

• Record Type: Journal Article
• Authors: Fan, Lili; Zhao, Hongwei; Zhao, Haoyu
• Is Part Of: IET image processing. Volume 14:Issue 15(2020)
• Page Start: 3802
• Page End: 3811
• Publication Date: 2021-02-18
• Subjects: optimisation -- learning (artificial intelligence) -- feature extraction -- data mining -- image retrieval -- image representation; embedding quality -- wise deep image embedding optimisation algorithm -- ranked list learning -- network optimisation -- fine‐grained image retrieval -- hard sample mining method Top‐k -- negative samples -- selected query sample -- ranking list -- positive samples -- optimisation rules -- instance weighting -- sample information -- optimisation direction -- end‐to‐end optimisation -- wise optimisation -- map images -- image retrieval performance -- informative pair weighting; Image processing -- Periodicals; 621.36705
• DOI: 10.1049/iet-ipr.2020.0454
• Languages: English
• ISSNs: 1751-9659

Abstract

Abstract : Deep image embedding learns how to map images onto feature vectors. Image retrieval performance is often used to evaluate embedding quality. In this study, the authors proposed a wise deep image embedding optimisation (WDIEO) algorithm based on informative pair weighting and ranked list learning (IPWRLL) for network optimisation of fine‐grained image retrieval. First, a hard sample mining method Top‐ k is proposed to select positive and negative samples. Then, for the selected query sample, a ranking list is obtained by comparing the similarity between samples in the data set and the query sample, and the sample is labelled according to the similarity. Finally, for positive samples, two optimisation rules with different functions are used, while ensuring two key issues of instance weighting and intra‐class data distribution. For negative samples, different from the widely adopted methods based on the weight of sample information, the authors' algorithm's weights are set according to the ranking list, which keeps the inter‐class data distribution and the optimisation direction consistent with the loss reduction direction. The WDIEO‐IPWRLL model is an end‐to‐end optimisation that can share parameters in the testing process. Experiments show that their proposed model achieves the state‐of‐the‐art performance on the benchmark data set.