conftrace_

2023 CVPR CVPR 2023

Learning a Depth Covariance Function

Abstract

We propose learning a depth covariance function with applications to geometric vision tasks. Given RGB images as input, the covariance function can be flexibly used to define priors over depth functions, predictive distributions given observations, and methods for active point selection. We leverage these techniques for a selection of downstream tasks: depth completion, bundle adjustment, and monocular dense visual odometry.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Computer Vision

Authors

Eric Dexheimer, Andrew J. Davison

Topics

Artificial Intelligence > Bayesian & Probabilistic > Probabilistic Modeling Machine Learning > Core Methods > Representation Learning Machine Learning > Optimization & Theory > Bayesian Inference Computer Vision > Analysis > Depth Estimation Machine Learning > Bayesian & Probabilistic > Gaussian Processes Computer Vision > Domain-Specific > Robotics

Keywords

visual odometry bayesian prior depth completion bundle adjustment geometric vision depth covariance function monocular dense visual odometry predictive depth distribution active point selection

More from CVPR 2023

All CVPR 2023 papers →

A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction From In-the-Wild Images

PersonNeRF: Personalized Reconstruction From Photo Collections

Enhanced Multimodal Representation Learning With Cross-Modal KD

Evading DeepFake Detectors via Adversarial Statistical Consistency

Referring Image Matting

V2V4Real: A Real-World Large-Scale Dataset for Vehicle-to-Vehicle Cooperative Perception