Research map / Label-efficient and robust learning / Recognition and long-tail
DA4AD: End-to-End Deep Attention-based Visual Localization for Autonomous Driving
Yao Zhou, Guowei Wan, Shenhua Hou, Li Yu, Gang Wang, Xiaofei Rui, Shiyu Song
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 weighted FPS (WFPS) instead of FPS with reduce average (ablation) | DA4AD · Apollo-DaoxiangLake localization (keypoint selection ablation) | horizontal RMS (m): 0.135 -> 0.058 | Improved |
| use weighted FPS (WFPS) instead of FPS with weighted average (ablation) | DA4AD · Apollo-DaoxiangLake localization (keypoint selection ablation) | horizontal RMS (m): 0.306 -> 0.063 | Improved |
| use weighted FPS (WFPS) instead of FPS with reduce average (ablation) | DA4AD · Apollo-DaoxiangLake localization (keypoint selection ablation) | N/A column (%): 98.1 -> 100.0 | Improved |
| use reduce average instead of weighted average in WFM with WFPS (ablation) | DA4AD · Apollo-DaoxiangLake localization (keypoint selection ablation) | horizontal RMS (m): 0.063 -> 0.058 | Improved |
| use reduce average instead of weighted average in WFM with WFPS (ablation) | DA4AD · Apollo-DaoxiangLake localization (keypoint selection ablation) | maximum horizontal error (m): 2.419 -> 2.617 | Hurt |
| use reduce average instead of weighted average in WFM with FPS (ablation) | DA4AD · Apollo-DaoxiangLake localization (keypoint selection ablation) | horizontal RMS (m): 0.306 -> 0.135 | Improved |
| use three cameras instead of single front camera | DA4AD · Apollo-DaoxiangLake localization | horizontal RMS (m): 0.123 -> 0.058 | Improved |
| use learned attention-based method instead of modified HF-Net (HFNet++) | DA4AD · Apollo-DaoxiangLake localization | horizontal RMS (m): 0.176 -> 0.058 | Improved |
| use learned HF-Net features instead of SIFT in HF-Net framework | DA4AD · Apollo-DaoxiangLake localization | horizontal RMS (m): 0.244 -> 0.176 | Improved |
| replace LiDAR localization with vision-based Ours | DA4AD · Apollo-DaoxiangLake localization | horizontal RMS (m): 0.053 -> 0.058 | Mixed |
Closest papers by meaning
- Deep Multi-Task Learning for Joint Localization, Perception, and Prediction (CVPR 2021 · similarity 0.88)
- From Coarse to Fine: Robust Hierarchical Localization at Large Scale (CVPR 2019 · similarity 0.87)
- L3-Net: Towards Learning Based LiDAR Localization for Autonomous Driving (CVPR 2019 · similarity 0.87)
- SFD2: Semantic-Guided Feature Detection and Description (CVPR 2023 · similarity 0.87)
- DALDet: Depth-Aware Learning Based Object Detection for Autonomous Driving (AAAI 2024 · similarity 0.85)
- DeepPointMap: Advancing LiDAR SLAM with Unified Neural Descriptors (AAAI 2024 · similarity 0.85)
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.
