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

In this paper, we introduce SalsaNext for the uncertainty-aware semanticsegmentation of a full 3D LiDAR point cloud in real-time. SalsaNext is the nextversion of SalsaNet [1] which has an encoder-decoder architecture where theencoder unit has a set of ResNet blocks and the decoder part combines upsampledfeatures from the residual blocks. In contrast to SalsaNet, we introduce a newcontext module, replace the ResNet encoder blocks with a new residual dilatedconvolution stack with gradually increasing receptive fields and add thepixel-shuffle layer in the decoder. Additionally, we switch from strideconvolution to average pooling and also apply central dropout treatment. Todirectly optimize the Jaccard index, we further combine the weightedcross-entropy loss with Lovasz-Softmax loss [2]. We finally inject a Bayesiantreatment to compute the epistemic and aleatoric uncertainties for each pointin the cloud. We provide a thorough quantitative evaluation on theSemantic-KITTI dataset [3], which demonstrates that the proposed SalsaNextoutperforms other state-of-the-art semantic segmentation networks and ranksfirst on the Semantic-KITTI leaderboard. We also release our source codehttps://github.com/TiagoCortinhal/SalsaNext.

SalsaNext: Fast, Uncertainty-aware Semantic Segmentation of LiDAR Point Clouds for Autonomous Driving