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

Transferring knowledge learned from the labeled source domain to the rawtarget domain for unsupervised domain adaptation (UDA) is essential to thescalable deployment of autonomous driving systems. State-of-the-art methods inUDA often employ a key idea: utilizing joint supervision signals from bothsource and target domains for self-training. In this work, we improve andextend this aspect. We present ConDA, a concatenation-based domain adaptationframework for LiDAR segmentation that: 1) constructs an intermediate domainconsisting of fine-grained interchange signals from both source and targetdomains without destabilizing the semantic coherency of objects and backgroundaround the ego-vehicle; and 2) utilizes the intermediate domain forself-training. To improve the network training on the source domain andself-training on the intermediate domain, we propose an anti-aliasingregularizer and an entropy aggregator to reduce the negative effect caused bythe aliasing artifacts and noisy pseudo labels. Through extensive studies, wedemonstrate that ConDA significantly outperforms prior arts in mitigatingdomain gaps.

ConDA: Unsupervised Domain Adaptation for LiDAR Segmentation via Regularized Domain Concatenation