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

In this work, we present a methodology to shape a fisheye-specificrepresentation space that reflects the interaction between distortion andsemantic context present in this data modality. Fisheye data has the widerfield of view advantage over other types of cameras, but this comes at theexpense of high radial distortion. As a result, objects further from the centerexhibit deformations that make it difficult for a model to identify theirsemantic context. While previous work has attempted architectural and trainingaugmentation changes to alleviate this effect, no work has attempted to guidethe model towards learning a representation space that reflects thisinteraction between distortion and semantic context inherent to fisheye data.We introduce an approach to exploit this relationship by first extractingdistortion class labels based on an object's distance from the center of theimage. We then shape a backbone's representation space with a weightedcontrastive loss that constrains objects of the same semantic class anddistortion class to be close to each other within a lower dimensional embeddingspace. This backbone trained with both semantic and distortion information isthen fine-tuned within an object detection setting to empirically evaluate thequality of the learnt representation. We show this method leads to performanceimprovements by as much as 1.1% mean average precision over standard objectdetection strategies and .6% improvement over other state of the artrepresentation learning approaches.

Exploiting the Distortion-Semantic Interaction in Fisheye Data