2023/04/18 by Zeeshan Khan Suri, Suri, Zeeshan Khan
Computer Science · #Advanced Image Processing Techniques #Advanced Vision and Imaging #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #FOS: Mathematics #Machine Learning (cs.LG) #Optical measurement and interference techniques #Optimization and Control (math.OC) #Robotics (cs.RO)
paper · pdf · doi:10.48550/arxiv.2304.08916
openalex publication_date 2023/04/18 · openalex created_date 2023/04/22 · openalex updated_date 2026/07/28
Self-supervised monocular depth estimation approaches suffer not only from scale ambiguity but also infer temporally inconsistent depth maps w.r.t. scale. While disambiguating scale during training is not possible without some kind of ground truth supervision, having scale consistent depth predictions would make it possible to calculate scale once during inference as a post-processing step and use it over-time. With this as a goal, a set of temporal consistency losses that minimize pose inconsistencies over time are introduced. Evaluations show that introducing these constraints not only reduces depth inconsistencies but also improves the baseline performance of depth and ego-motion prediction.