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 exploitation of visible spectrum datasets has led deep networks to showremarkable success. However, real-world tasks include low-lighting conditionswhich arise performance bottlenecks for models trained on large-scale RGB imagedatasets. Thermal IR cameras are more robust against such conditions.Therefore, the usage of thermal imagery in real-world applications can beuseful. Unsupervised domain adaptation (UDA) allows transferring informationfrom a source domain to a fully unlabeled target domain. Despite substantialimprovements in UDA, the performance gap between UDA and its supervisedlearning counterpart remains significant. By picking a small number of targetsamples to annotate and using them in training, active domain adaptation triesto mitigate this gap with minimum annotation expense. We propose an activedomain adaptation method in order to examine the efficiency of combining thevisible spectrum and thermal imagery modalities. When the domain gap isconsiderably large as in the visible-to-thermal task, we may conclude that themethods without explicit domain alignment cannot achieve their full potential.To this end, we propose a spectral transfer guided active domain adaptationmethod to select the most informative unlabeled target samples while aligningsource and target domains. We used the large-scale visible spectrum datasetMS-COCO as the source domain and the thermal dataset FLIR ADAS as the targetdomain to present the results of our method. Extensive experimental evaluationdemonstrates that our proposed method outperforms the state-of-the-art activedomain adaptation methods. The code and models are publicly available.

Spectral Transfer Guided Active Domain Adaptation For Thermal Imagery