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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recently, there have been a plethora of classification and detection systemsfrom RGB as well as 3D images. In this work, we describe a new 3D objectdetection system from an RGB-D or depth-only point cloud. Our system firstdetects objects in 2D (either RGB or pseudo-RGB constructed from depth). Thenext step is to detect 3D objects within the 3D frustums these 2D detectionsdefine. This is achieved by voxelizing parts of the frustums (since frustumscan be really large), instead of using the whole frustums as done in earlierwork. The main novelty of our system has to do with determining which parts (3Dproposals) of the frustums to voxelize, thus allowing us to provide highresolution representations around the objects of interest. It also allows oursystem to have reduced memory requirements. These 3D proposals are fed to anefficient ResNet-based 3D Fully Convolutional Network (FCN). Our 3D detectionsystem is fast and can be integrated into a robotics platform. With respect tosystems that do not perform voxelization (such as PointNet), our methods canoperate without the requirement of subsampling of the datasets. We have alsointroduced a pipelining approach that further improves the efficiency of oursystem. Results on SUN RGB-D dataset show that our system, which is based on asmall network, can process 20 frames per second with comparable detectionresults to the state-of-the-art, achieving a 2 times speedup.

Frustum VoxNet for 3D object detection from RGB-D or Depth images