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

Global feature based Pedestrian Attribute Recognition (PAR) models are oftenpoorly localized when using Grad-CAM for attribute response analysis, which hasa significant impact on the interpretability, generalizability and performance.Previous researches have attempted to improve generalization and interpretationthrough meticulous model design, yet they often have neglected or underutilizedeffective prior information crucial for PAR. To this end, a novel Scale andSpatial Priors Guided Network (SSPNet) is proposed for PAR, which is mainlycomposed of the Adaptive Feature Scale Selection (AFSS) and Prior LocationExtraction (PLE) modules. The AFSS module learns to provide reasonable scaleprior information for different attribute groups, allowing the model to focuson different levels of feature maps with varying semantic granularity. The PLEmodule reveals potential attribute spatial prior information, which avoidsunnecessary attention on irrelevant areas and lowers the risk of modelover-fitting. More specifically, the scale prior in AFSS is adaptively learnedfrom different layers of feature pyramid with maximum accuracy, while thespatial priors in PLE can be revealed from part feature with differentgranularity (such as image blocks, human pose keypoint and sparse samplingpoints). Besides, a novel IoU based attribute localization metric is proposedfor Weakly-supervised Pedestrian Attribute Localization (WPAL) based on theimproved Grad-CAM for attribute response mask. The experimental results on theintra-dataset and cross-dataset evaluations demonstrate the effectiveness ofour proposed method in terms of mean accuracy (mA). Furthermore, it alsoachieves superior performance on the PCS dataset for attribute localization interms of IoU. Code will be released at https://github.com/guotengg/SSPNet.

SSPNet: Scale and Spatial Priors Guided Generalizable and Interpretable Pedestrian Attribute Recognition