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

3D object detection is a significant task for autonomous driving. Recentlywith the progress of vision transformers, the 2D object detection problem isbeing treated with the set-to-set loss. Inspired by these approaches on 2Dobject detection and an approach for multi-view 3D object detection DETR3D, wepropose MSF3DDETR: Multi-Sensor Fusion 3D Detection Transformer architecture tofuse image and LiDAR features to improve the detection accuracy. Our end-to-endsingle-stage, anchor-free and NMS-free network takes in multi-view images andLiDAR point clouds and predicts 3D bounding boxes. Firstly, we link the objectqueries learnt from data to the image and LiDAR features using a novelMSF3DDETR cross-attention block. Secondly, the object queries interacts witheach other in multi-head self-attention block. Finally, MSF3DDETR block isrepeated for $L$ number of times to refine the object queries. The MSF3DDETRnetwork is trained end-to-end on the nuScenes dataset using Hungarian algorithmbased bipartite matching and set-to-set loss inspired by DETR. We present bothquantitative and qualitative results which are competitive to thestate-of-the-art approaches.

MSF3DDETR: Multi-Sensor Fusion 3D Detection Transformer for Autonomous Driving