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

Multi-sensor fusion is essential for accurate 3D object detection inself-driving systems. Camera and LiDAR are the most commonly used sensors, andusually, their fusion happens at the early or late stages of 3D detectors withthe help of regions of interest (RoIs). On the other hand, fusion at theintermediate level is more adaptive because it does not need RoIs frommodalities but is complex as the features of both modalities are presented fromdifferent points of view. In this paper, we propose a new intermediate-levelmulti-modal fusion (mmFUSION) approach to overcome these challenges. First, themmFUSION uses separate encoders for each modality to compute features at adesired lower space volume. Second, these features are fused throughcross-modality and multi-modality attention mechanisms proposed in mmFUSION.The mmFUSION framework preserves multi-modal information and learns tocomplement modalities' deficiencies through attention weights. The strongmulti-modal features from the mmFUSION framework are fed to a simple 3Ddetection head for 3D predictions. We evaluate mmFUSION on the KITTI andNuScenes dataset where it performs better than available early, intermediate,late, and even two-stage based fusion schemes. The code with the mmdetection3Dproject plugin will be publicly available soon.

mmFUSION: Multimodal Fusion for 3D Objects Detection