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

Robustly classifying ground infrastructure such as roads and street crossingsis an essential task for mobile robots operating alongside pedestrians. Whilemany semantic segmentation datasets are available for autonomous vehicles,models trained on such datasets exhibit a large domain gap when deployed onrobots operating in pedestrian spaces. Manually annotating images recorded frompedestrian viewpoints is both expensive and time-consuming. To overcome thischallenge, we propose TrackletMapper, a framework for annotating ground surfacetypes such as sidewalks, roads, and street crossings from object trackletswithout requiring human-annotated data. To this end, we project the robotego-trajectory and the paths of other traffic participants into the ego-viewcamera images, creating sparse semantic annotations for multiple types ofground surfaces from which a ground segmentation model can be trained. Wefurther show that the model can be self-distilled for additional performancebenefits by aggregating a ground surface map and projecting it into the cameraimages, creating a denser set of training annotations compared to the sparsetracklet annotations. We qualitatively and quantitatively attest our findingson a novel large-scale dataset for mobile robots operating in pedestrian areas.Code and dataset will be made available athttp://trackletmapper.cs.uni-freiburg.de.

TrackletMapper: Ground Surface Segmentation and Mapping from Traffic Participant Trajectories