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

3D LiDAR sensors are indispensable for the robust vision of autonomous mobilerobots. However, deploying LiDAR-based perception algorithms often fails due toa domain gap from the training environment, such as inconsistent angularresolution and missing properties. Existing studies have tackled the issue bylearning inter-domain mapping, while the transferability is constrained by thetraining configuration and the training is susceptible to peculiar lossy noisescalled ray-drop. To address the issue, this paper proposes a generative modelof LiDAR range images applicable to the data-level domain transfer. Motivatedby the fact that LiDAR measurement is based on point-by-point range imaging, wetrain an implicit image representation-based generative adversarial networksalong with a differentiable ray-drop effect. We demonstrate the fidelity anddiversity of our model in comparison with the point-based and image-basedstate-of-the-art generative models. We also showcase upsampling and restorationapplications. Furthermore, we introduce a Sim2Real application for LiDARsemantic segmentation. We demonstrate that our method is effective as arealistic ray-drop simulator and outperforms state-of-the-art methods.

Generative Range Imaging for Learning Scene Priors of 3D LiDAR Data