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

The prevalence of smartphone and consumer camera has led to more evidence inthe form of digital images, which are mostly taken in uncontrolled anduncooperative environments. In these images, criminals likely hide or covertheir faces while their hands are observable in some cases, creating achallenging use case for forensic investigation. Many existing hand-basedrecognition methods perform well for hand images collected in controlledenvironments with user cooperation. However, their performance deterioratessignificantly in uncontrolled and uncooperative environments. A recent work hasexposed the potential of hand recognition in these environments. However, onlythe palmar regions were considered, and the recognition performance is stillfar from satisfactory. To improve the recognition accuracy, an algorithmintegrating a multi-spatial transformer network (MSTN) and multiple lossfunctions is proposed to fully utilize information in full hand images. MSTN isfirstly employed to localize the palms and fingers and estimate the alignmentparameters. Then, the aligned images are further fed into pretrainedconvolutional neural networks, where features are extracted. Finally, atraining scheme with multiple loss functions is used to train the networkend-to-end. To demonstrate the effectiveness of the proposed algorithm, thetrained model is evaluated on NTU-PI-v1 database and six benchmark databasesfrom different domains. Experimental results show that the proposed algorithmperforms significantly better than the existing methods in these uncontrolledand uncooperative environments and has good generalization capabilities tosamples from different domains.

Improving Hand Recognition in Uncontrolled and Uncooperative Environments using Multiple Spatial Transformers and Loss Functions