Research map / 3D vision / Depth and stereo
Improving 360 Monocular Depth Estimation via Non-local Dense Prediction Transformer and Joint Supervised and Self-Supervised Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use Ours w/ NLFB instead of HoHoNet | 360 monocular depth estimation · Stanford3D | AbsRel: 0.0901 -> 0.0649 | Helped |
| Use Ours w/ NLFB instead of HoHoNet | 360 monocular depth estimation · Matterport3D | AbsRel: 0.0671 -> 0.0700 | Hurt |
| Use Ours w/ NLFB instead of HoHoNet | 360 monocular depth estimation · SunCG | AbsRel: 0.0827 -> 0.0715 | Helped |
| Use Ours w/ NLFB instead of HoHoNet | 360 monocular depth estimation · Stanford | AbsRel: 0.0801 -> 0.0666 | Helped |
| Add depth consistency to self-supervised image consistency (ablation) | Self-supervised depth estimation · Structure3D | AbsRel: 0.1056 -> 0.1048 | Helped |
| Add image consistency to supervised pixel-wise and gradient losses (ablation) | 360 monocular depth estimation · Structure3D | AbsRel: 0.0867 -> 0.0840 | Helped |
| Add depth consistency to supervised losses and image consistency (ablation) | Joint supervised and self-supervised depth estimation · Structure3D | AbsRel: 0.0840 -> 0.0802 | Helped |
| Add self-supervised image and depth consistency to supervised losses (ablation) | 360 monocular depth estimation · Structure3D | AbsRel: 0.0867 -> 0.0802 | Helped |
| Add NLFB to joint learning (ablation) | Joint supervised and self-supervised depth estimation · Structure3D | AbsRel: 0.0802 -> 0.0781 | Helped |
| Remove fine-tuning from the rectilinear-depth pretrained model (ablation) | Non-local DPT · Structure3D | AbsRel: 0.0781 -> 1.000 | Hurt |
| Remove robust depth-scale adjustment (ablation) | Joint supervised and self-supervised depth estimation · Structure3D | AbsRel: 0.0781 -> 0.3793 | Hurt |
Closest papers by meaning
- Self-Supervised Monocular Trained Depth Estimation Using Self-Attention and Discrete Disparity Volume (CVPR 2020 · similarity 0.91)
- DME: Unveiling the Bias for Better Generalized Monocular Depth Estimation (AAAI 2024 · similarity 0.87)
- GeoDepth: From Point-to-Depth to Plane-to-Depth Modeling for Self-Supervised Monocular Depth Estimation (CVPR 2025 · similarity 0.87)
- Improved Monocular Depth Prediction Using Distance Transform Over Pre-semantic Contours with Self-supervised Neural Networks (CVPR 2025 · similarity 0.87)
- Mining Supervision for Dynamic Regions in Self-Supervised Monocular Depth Estimation (CVPR 2024 · similarity 0.87)
- Depth Anywhere: Enhancing 360 Monocular Depth Estimation via Perspective Distillation and Unlabeled Data Augmentation (NeurIPS 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.