Research map / 3D vision / Depth and stereo
Seeing Depth Through Frequency and Motion: A Progressive Training Paradigm for Monocular Depth Estimation
Ke Li, Bolin Song, Hongbo Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 FGDepth (Frequency-Guided Sampling) to the depth network under joint training (ablation) | baseline depth network · KITTI Eigen split | Sq Rel: 0.727 -> 0.712 | Improved |
| add PQNet as pose network under joint training (ablation) | baseline · KITTI Eigen split | Sq Rel: 0.727 -> 0.723 | Improved |
| combine FGDepth and PQNet under joint training (ablation) | baseline · KITTI Eigen split | Sq Rel: 0.727 -> 0.704 | Improved |
| use progressive three-stage training instead of single-stage joint training (ablation) | FGDepth + PQNet · KITTI Eigen split | Sq Rel: 0.704 -> 0.679 | Improved |
| alternate depth and pose updates every mini-batch instead of three-stage training (ablation) | FGDepth + PQNet · KITTI Eigen split | Abs Rel: 0.098 (joint) -> 0.117 | Hurt |
| alternate depth and pose updates every epoch instead of joint training (ablation) | FGDepth + PQNet · KITTI Eigen split | Sq Rel: 0.704 (joint) -> 0.710 | No clear effect |
| add PQNet to FGDepth under joint training (ablation) | pose estimation · KITTI Odometry Seq. 09 | ATE: 0.019 -> 0.018 | Improved |
| use three-stage training instead of joint training (ablation) | FGDepth + PQNet · KITTI Odometry Seq. 09 | ATE: 0.018 -> 0.016 | Improved |
| use three-stage training instead of joint training (ablation) | FGDepth + PQNet · KITTI Odometry Seq. 10 | ATE: 0.015 -> 0.014 | No clear effect |
| use FGDepth instead of MonoViT | KITTI depth model · KITTI Eigen split | Sq Rel: 0.708 -> 0.679 | Improved |
| use FGDepth instead of MonoViT in zero-shot transfer | KITTI-trained depth model · Make3D (zero-shot) | Sq Rel: 2.758 -> 2.627 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- GeoDepth — GeoDepth: From Point-to-Depth to Plane-to-Depth Modeling for Self-Supervised Monocular Depth Estimation (CVPR 2025) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Digging Into Self-Supervised Monocular Depth Estimation (ICCV 2019 · similarity 0.90)
- The Temporal Opportunist: Self-Supervised Multi-Frame Monocular Depth (CVPR 2021 · similarity 0.90)
- Robust Consistent Video Depth Estimation (CVPR 2021 · similarity 0.89)
- Mining Supervision for Dynamic Regions in Self-Supervised Monocular Depth Estimation (CVPR 2024 · similarity 0.88)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.88)
- DepthFM: Fast Generative Monocular Depth Estimation with Flow Matching (AAAI 2025 · similarity 0.88)
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.
