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 the domain of autonomous driving, deep learning has substantially improvedthe 3D object detection accuracy for LiDAR and stereo camera data alike. Whiledeep networks are great at generalization, they are also notorious to over-fitto all kinds of spurious artifacts, such as brightness, car sizes and models,that may appear consistently throughout the data. In fact, most datasets forautonomous driving are collected within a narrow subset of cities within onecountry, typically under similar weather conditions. In this paper we considerthe task of adapting 3D object detectors from one dataset to another. Weobserve that naively, this appears to be a very challenging task, resulting indrastic drops in accuracy levels. We provide extensive experiments toinvestigate the true adaptation challenges and arrive at a surprisingconclusion: the primary adaptation hurdle to overcome are differences in carsizes across geographic areas. A simple correction based on the average carsize yields a strong correction of the adaptation gap. Our proposed method issimple and easily incorporated into most 3D object detection frameworks. Itprovides a first baseline for 3D object detection adaptation across countries,and gives hope that the underlying problem may be more within grasp than onemay have hoped to believe. Our code is available athttps://github.com/cxy1997/3D_adapt_auto_driving.

Train in Germany, Test in The USA: Making 3D Object Detectors Generalize