Research map / 3D vision / Depth and stereo
Unsupervised Learning of Monocular Depth Estimation and Visual Odometry With Deep Feature Reconstruction
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 |
|---|---|---|---|
| add temporal pairs to stereo training (ablation) | Stereo depth model · KITTI Eigen Split | Abs Rel: 0.143 -> 0.135 | Helped |
| add ImageNet conv1 feature reconstruction loss (ablation) | Stereo depth model · KITTI Eigen Split | Abs Rel: 0.143 -> 0.136 | Helped |
| add self-supervised KITTI descriptor feature loss (ablation) | Stereo depth model · KITTI Eigen Split | Abs Rel: 0.143 -> 0.130 | Helped |
| add NYUv2 descriptor feature loss (ablation) | Stereo depth model · KITTI Eigen Split | Abs Rel: 0.143 -> 0.132 | Helped |
| add temporal pairs and NYUv2 feature loss (ablation) | Stereo depth model · KITTI Eigen Split | Abs Rel: 0.143 -> 0.128 | Helped |
| add temporal pairs to stereo training (ablation) | Learnable-upsampler depth model · KITTI Eigen Split | Abs Rel: 0.155 -> 0.141 | Helped |
| add depth-decoder self-embedded feature reconstruction loss (ablation) | Learnable-upsampler depth model · KITTI Eigen Split | Abs Rel: 0.155 -> 0.142 | Helped |
| add temporal pairs and depth-decoder feature loss (ablation) | Learnable-upsampler depth model · KITTI Eigen Split | Abs Rel: 0.155 -> 0.137 | Helped |
| use stereo-video training instead of monocular training | Depth and visual odometry framework · KITTI Eigen Split, 80m cap | Abs Rel: 0.208 -> 0.144 | Helped |
| use stereo-based rather than monocular-based training | Visual odometry model · KITTI Odometry Seq. 09 | translational drift error (%): 17.84 -> 11.93 | Helped |
| integrate frame-to-frame poses without bundle adjustment | Ours (Temporal) · KITTI Odometry Seq. 10 | translational drift error (%): 3.68 vs 12.45 | Hurt |
Closest papers by meaning
- D3VO: Deep Depth, Deep Pose and Deep Uncertainty for Monocular Visual Odometry (CVPR 2020 · similarity 0.92)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.90)
- ZeroVO: Visual Odometry with Minimal Assumptions (CVPR 2025 · similarity 0.87)
- Leveraging Consistent Spatio-Temporal Correspondence for Robust Visual Odometry (AAAI 2025 · similarity 0.87)
- Adaptive Fusion of Single-View and Multi-View Depth for Autonomous Driving (CVPR 2024 · similarity 0.87)
- Seurat: From Moving Points to Depth (CVPR 2025 · 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.