Discriminative Multi-View Privileged Information Learning for Image Re-Ranking

Abstract

Conventional multi-view re-ranking methods usually perform asymmetrical matching between the region of interest (ROI) in the query image and the whole target image for similarity computation. Due to the inconsistency in the visual appearance, this practice tends to degrade the retrieval accuracy particularly when the image ROI, which is usually interpreted as the image objectness, accounts for a smaller region in the image. Since Privileged Information (PI), which can be viewed as the image prior, enables well characterizing the image objectness, we are aiming at leveraging PI for further improving the performance of the multi-view re-ranking accuracy in this paper. Towards this end, we propose a discriminative multi-view re-ranking approach in which both the original global image visual contents and the local auxiliary PI features are simultaneously integrated into a unified training framework for generating the latent subspaces with sufficient discriminating power. For the on-thefly re-ranking, since the multi-view PI features are unavailable, we only project the original multi-view image representations onto the latent subspace, and thus the re-ranking can be achieved by computing and sorting the distances from the multi-view embeddings to the separating hyperplane. Extensive experimental evaluations on the two public benchmarks Oxford5k and Paris6k reveal our approach provides further performance boost for accurate image re-ranking, whilst the comparative study demonstrates the advantage of our method against other multi-view re-ranking methods.

DMvLNet: deep multiview learning network for blindly accessing image quality

“…It is well known that multiview learning allows for uncovering the mutual correlation among different views of data or features. 25,26 In particular, it can fully take into account the complementarity among different views for exploring the intrinsic structure underlying the original multiview feature spaces. Thus, multiview learning is extensively applied to a wide range of vision tasks.…”

Determination of multi-modal dimension metric learning with application to web image retrieval

“…The estimation of the distance or similarity between objects in many machine learning and pattern recognition algorithms requires a measure. Content-Based Information Retrieval (CBIR) systems, for example, use a similarity function to rank items based on their resemblance to a query picture [1] [2]. The purpose of CBIR is to search photographs by assessing the image's actual content rather than information such as keywords, title, and author, thus a great deal of work has been devoted to the exploration of various low-level feature descriptors for picture representation [2].…”

Robust learning with imperfect privileged information

“…Recently, LUPI has attracted more and more attentions in different applications and been investigated in many existing algorithms [13,14,15,16,17].…”