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

Accurate per-object distance estimation is crucial in safety-criticalapplications such as autonomous driving, surveillance, and robotics. Existingapproaches rely on two scales: local information (i.e., the bounding boxproportions) or global information, which encodes the semantics of the scene aswell as the spatial relations with neighboring objects. However, theseapproaches may struggle with long-range objects and in the presence of strongocclusions or unusual visual patterns. In this respect, our work aims tostrengthen both local and global cues. Our architecture -- named DistFormer --builds upon three major components acting jointly: i) a robust context encoderextracting fine-grained per-object representations; ii) a maskedencoder-decoder module exploiting self-supervision to promote the learning ofuseful per-object features; iii) a global refinement module that aggregatesobject representations and computes a joint, spatially-consistent estimation.To evaluate the effectiveness of DistFormer, we conduct experiments on thestandard KITTI dataset and the large-scale NuScenes and MOTSynth datasets. Suchdatasets cover various indoor/outdoor environments, changing weatherconditions, appearances, and camera viewpoints. Our comprehensive analysisshows that DistFormer outperforms existing methods. Moreover, we further delveinto its generalization capabilities, showing its regularization benefits inzero-shot synth-to-real transfer.

DistFormer: Enhancing Local and Global Features for Monocular Per-Object Distance Estimation