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

Monocular depth predictors are typically trained on large-scale training setswhich are naturally biased w.r.t the distribution of camera poses. As a result,trained predictors fail to make reliable depth predictions for testing examplescaptured under uncommon camera poses. To address this issue, we propose twonovel techniques that exploit the camera pose during training and prediction.First, we introduce a simple perspective-aware data augmentation thatsynthesizes new training examples with more diverse views by perturbing theexisting ones in a geometrically consistent manner. Second, we propose aconditional model that exploits the per-image camera pose as prior knowledge byencoding it as a part of the input. We show that jointly applying the twomethods improves depth prediction on images captured under uncommon and evennever-before-seen camera poses. We show that our methods improve performancewhen applied to a range of different predictor architectures. Lastly, we showthat explicitly encoding the camera pose distribution improves thegeneralization performance of a synthetically trained depth predictor whenevaluated on real images.

Camera Pose Matters: Improving Depth Prediction by Mitigating Pose Distribution Bias