Research map / 3D vision / Depth and stereo
Radar-Guided Polynomial Fitting for Metric Depth Estimation
Patrick Rim, Hyoungseob Park, Vadim Ezhov, Jeffrey Moon, Alex Wong
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 |
|---|---|---|---|
| use polynomial fitting of degree 8 instead of scale-and-shift (degree 1) (ablation) | POLAR · nuScenes (80m) | MAE: 2156.8 -> 1407.8 | Improved |
| use polynomial fitting of degree 8 instead of scale-and-shift (ablation) | POLAR · ZJU-4DRadarCam (80m) | MAE: 1078.2 -> 629.6 | Improved |
| raise polynomial degree from 8 to 10 (ablation) | POLAR · nuScenes (80m) | MAE: 1407.8 -> 1463.7 | Hurt |
| add monotonicity (first-derivative) loss (ablation) | POLAR · nuScenes (80m) | MAE: 1921.1 -> 1407.8 | Improved |
| add monotonicity loss (ablation) | POLAR · View-of-Delft (80m) | MAE: 1924.5 -> 1500.1 | Improved |
| add learnable prototypes for radar aggregation instead of self-attention (ablation) | POLAR · nuScenes (80m) | MAE: 1615.5 -> 1407.8 | Improved |
| use feature aggregation (Eq. 2) instead of cross-attention (ablation) | POLAR · nuScenes (80m) | MAE: 1454.1 -> 1407.8 | Improved |
| add cross-modality attention fusion (ablation) | POLAR · nuScenes (80m) | MAE: 2238.8 -> 1407.8 | Improved |
| predict polynomial coefficients instead of direct per-pixel depth decoding (ablation) | POLAR · nuScenes (80m) | MAE: 1968.0 -> 1407.8 | Improved |
| POLAR vs TacoDepth on nuScenes 50m | radar-camera depth · nuScenes (50m) | MAE: 1046.8 -> 1014.4 | 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.
- TacoDepth — TacoDepth: Towards Efficient Radar-Camera Depth Estimation with One-stage Fusion (CVPR 2025) · better on 8 of 8 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
- Selection, Not Fusion: Radar-Modulated State Space Models for Radar-Camera Depth Estimation (NeurIPS 2026 · similarity 0.88)
- TacoDepth: Towards Efficient Radar-Camera Depth Estimation with One-stage Fusion (CVPR 2025 · similarity 0.87)
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.87)
- RSA: Resolving Scale Ambiguities in Monocular Depth Estimators through Language Descriptions (NeurIPS 2024 · similarity 0.86)
- SharpDepth: Sharpening Metric Depth Predictions Using Diffusion Distillation (CVPR 2025 · similarity 0.86)
- Flexible Depth Completion for Sparse and Varying Point Densities (CVPR 2024 · similarity 0.86)
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.
