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

Solving depth estimation with monocular cameras enables the possibility ofwidespread use of cameras as low-cost depth estimation sensors in applicationssuch as autonomous driving and robotics. However, learning such a scalabledepth estimation model would require a lot of labeled data which is expensiveto collect. There are two popular existing approaches which do not requireannotated depth maps: (i) using labeled synthetic and unlabeled real data in anadversarial framework to predict more accurate depth, and (ii) unsupervisedmodels which exploit geometric structure across space and time in monocularvideo frames. Ideally, we would like to leverage features provided by bothapproaches as they complement each other; however, existing methods do notadequately exploit these additive benefits. We present $S^3$Net, aself-supervised framework which combines these complementary features: we usesynthetic and real-world images for training while exploiting geometric,temporal, as well as semantic constraints. Our novel consolidated architectureprovides a new state-of-the-art in self-supervised depth estimation usingmonocular videos. We present a unique way to train this self-supervisedframework, and achieve (i) more than $15\%$ improvement over previous syntheticsupervised approaches that use domain adaptation and (ii) more than $10\%$improvement over previous self-supervised approaches which exploit geometricconstraints from the real data.

$S^3$Net: Semantic-Aware Self-supervised Depth Estimation with Monocular Videos and Synthetic Data