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

Many LiDAR-based methods for detecting large objects, single-class objectdetection, or under easy situations were claimed to perform quite well.However, their performances of detecting small objects or under hard situationsdid not surpass those of the fusion-based ones due to failure to leverage theimage semantics. In order to elevate the detection performance in a complicatedenvironment, this paper proposes a deep learning (DL)-embedded fusion-basedmulti-class 3D object detection network which admits both LiDAR and camerasensor data streams, named Voxel-Pixel Fusion Network (VPFNet). Inside thisnetwork, a key novel component is called Voxel-Pixel Fusion (VPF) layer, whichtakes advantage of the geometric relation of a voxel-pixel pair and fuses thevoxel features and the pixel features with proper mechanisms. Moreover, severalparameters are particularly designed to guide and enhance the fusion effectafter considering the characteristics of a voxel-pixel pair. Finally, theproposed method is evaluated on the KITTI benchmark for multi-class 3D objectdetection task under multilevel difficulty, and is shown to outperform allstate-of-the-art methods in mean average precision (mAP). It is also noteworthythat our approach here ranks the first on the KITTI leaderboard for thechallenging pedestrian class.

VPFNet: Voxel-Pixel Fusion Network for Multi-class 3D Object Detection