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

Although DETR-based 3D detectors can simplify the detection pipeline andachieve direct sparse predictions, their performance still lags behind densedetectors with post-processing for 3D object detection from point clouds. DETRsusually adopt a larger number of queries than GTs (e.g., 300 queries v.s. 40objects in Waymo) in a scene, which inevitably incur many false positivesduring inference. In this paper, we propose a simple yet effective sparse 3Ddetector, named Query Contrast Voxel-DETR (ConQueR), to eliminate thechallenging false positives, and achieve more accurate and sparser predictions.We observe that most false positives are highly overlapping in local regions,caused by the lack of explicit supervision to discriminate locally similarqueries. We thus propose a Query Contrast mechanism to explicitly enhancequeries towards their best-matched GTs over all unmatched query predictions.This is achieved by the construction of positive and negative GT-query pairsfor each GT, and a contrastive loss to enhance positive GT-query pairs againstnegative ones based on feature similarities. ConQueR closes the gap of sparseand dense 3D detectors, and reduces up to ~60% false positives. Oursingle-frame ConQueR achieves new state-of-the-art (sota) 71.6 mAPH/L2 on thechallenging Waymo Open Dataset validation set, outperforming previous sotamethods (e.g., PV-RCNN++) by over 2.0 mAPH/L2.

ConQueR: Query Contrast Voxel-DETR for 3D Object Detection