English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Visible-infrared person re-identification (VI-ReID), which aims to searchidentities across different spectra, is a challenging task due to largecross-modality discrepancy between visible and infrared images. The key toreduce the discrepancy is to filter out identity-irrelevant interference andeffectively learn modality-invariant person representations. In this paper, wepropose a novel Modality Restitution and Compensation Network (MRCN) to narrowthe gap between the two modalities. Specifically, we first reduce the modalitydiscrepancy by using two Instance Normalization (IN) layers. Next, to reducethe influence of IN layers on removing discriminative information and to reducemodality differences, we propose a Modality Restitution Module (MRM) and aModality Compensation Module (MCM) to respectively distill modality-irrelevantand modality-relevant features from the removed information. Then, themodality-irrelevant features are used to restitute to the normalized visibleand infrared features, while the modality-relevant features are used tocompensate for the features of the other modality. Furthermore, to betterdisentangle the modality-relevant features and the modality-irrelevantfeatures, we propose a novel Center-Quadruplet Causal (CQC) loss to encouragethe network to effectively learn the modality-relevant features and themodality-irrelevant features. Extensive experiments are conducted to validatethe superiority of our method on the challenging SYSU-MM01 and RegDB datasets.More remarkably, our method achieves 95.1% in terms of Rank-1 and 89.2% interms of mAP on the RegDB dataset.

MRCN: A Novel Modality Restitution and Compensation Network for Visible-Infrared Person Re-identification