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

Stable imaging in adverse environments (e.g., total darkness) makes thermalinfrared (TIR) cameras a prevalent option for night scene perception. However,the low contrast and lack of chromaticity of TIR images are detrimental tohuman interpretation and subsequent deployment of RGB-based vision algorithms.Therefore, it makes sense to colorize the nighttime TIR images by translatingthem into the corresponding daytime color images (NTIR2DC). Despite theimpressive progress made in the NTIR2DC task, how to improve the translationperformance of small object classes is under-explored. To address this problem,we propose a generative adversarial network incorporating feedback-based objectappearance learning (FoalGAN). Specifically, an occlusion-aware mixup moduleand corresponding appearance consistency loss are proposed to reduce thecontext dependence of object translation. As a representative example of smallobjects in nighttime street scenes, we illustrate how to enhance the realism oftraffic light by designing a traffic light appearance loss. To further improvethe appearance learning of small objects, we devise a dual feedback learningstrategy to selectively adjust the learning frequency of different samples. Inaddition, we provide pixel-level annotation for a subset of the Brno dataset,which can facilitate the research of NTIR image understanding under multipleweather conditions. Extensive experiments illustrate that the proposed FoalGANis not only effective for appearance learning of small objects, but alsooutperforms other image translation methods in terms of semantic preservationand edge consistency for the NTIR2DC task.

Nighttime Thermal Infrared Image Colorization with Feedback-based Object Appearance Learning