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

Vision-based depth estimation is a key feature in autonomous systems, whichoften relies on a single camera or several independent ones. In such amonocular setup, dense depth is obtained with either additional input from oneor several expensive LiDARs, e.g., with 64 beams, or camera-only methods, whichsuffer from scale-ambiguity and infinite-depth problems. In this paper, wepropose a new alternative of densely estimating metric depth by combining amonocular camera with a light-weight LiDAR, e.g., with 4 beams, typical oftoday's automotive-grade mass-produced laser scanners. Inspired by recentself-supervised methods, we introduce a novel framework, called LiDARTouch, toestimate dense depth maps from monocular images with the help of ``touches'' ofLiDAR, i.e., without the need for dense ground-truth depth. In our setup, theminimal LiDAR input contributes on three different levels: as an additionalmodel's input, in a self-supervised LiDAR reconstruction objective function,and to estimate changes of pose (a key component of self-supervised depthestimation architectures). Our LiDARTouch framework achieves new state of theart in self-supervised depth estimation on the KITTI dataset, thus supportingour choices of integrating the very sparse LiDAR signal with other visualfeatures. Moreover, we show that the use of a few-beam LiDAR alleviates scaleambiguity and infinite-depth issues that camera-only methods suffer from. Wealso demonstrate that methods from the fully-supervised depth-completionliterature can be adapted to a self-supervised regime with a minimal LiDARsignal.

LiDARTouch: Monocular metric depth estimation with a few-beam LiDAR