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

One essential feature of an autonomous train is minimizing collision riskswith third-party objects. To estimate the risk, the control system mustidentify topological information of all the rail routes ahead on which thetrain can possibly move, especially within merging or diverging rails. Thisway, the train can figure out the status of potential obstacles with respect toits route and hence, make a timely decision. Numerous studies have successfullyextracted all rail tracks as a whole within forward-looking images withoutconsidering element instances. Still, some image-based methods have employedhard-coded prior knowledge of railway geometry on 3D data to associateleft-right rails and generate rail route instances. However, we propose a railpath extraction pipeline in which left-right rail pixels of each rail routeinstance are extracted and associated through a fully convolutionalencoder-decoder architecture called TPE-Net. Two different regression branchesfor TPE-Net are proposed to regress the locations of center points of each railroute, along with their corresponding left-right pixels. Extracted rail pixelsare then spatially clustered to generate topological information of all thepossible train routes (ego-paths), discarding non-ego-path ones. Experimentalresults on a challenging, publicly released benchmark show true-positive-pixellevel average precision and recall of 0.9207 and 0.8721, respectively, at about12 frames per second. Even though our evaluation results are not higher thanthe SOTA, the proposed regression pipeline performs remarkably in extractingthe correspondences by looking once at the image. It generates strong railroute hypotheses without reliance on camera parameters, 3D data, andgeometrical constraints.

TPE-Net: Track Point Extraction and Association Network for Rail Path Proposal Generation