Research map / Detection and segmentation / CNN features
Towards End-to-End License Plate Detection and Recognition: A Large Dataset and Baseline
Zhenbo Xu, Wei Yang, Ajin Meng, Nanxue Lu, Huan Huang, Changchun Ying, Liusheng Huang
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 |
|---|---|---|---|
| extract recognition features from a higher layer (RP5/RP135) instead of only C1 (RP1) (ablation) | RPnet · CCPD-Base recognition | accuracy (%): 93.5 -> 97.5 | Improved |
| add lower layers C1 and C3 to a single C5 feature layer (RP5 to RP135) (ablation) | RPnet · CCPD-Base recognition | accuracy (%) | Improved |
| add layers C2 and C4 to RP135 (RP12345 using five layers) (ablation) | RPnet · CCPD-Base recognition | accuracy and FPS: 61 FPS -> 46 FPS | Hurt |
| share feature maps between detection and recognition (RPnet vs SSD300 + Holistic-CNN) | RPnet · CCPD recognition | FPS: 35 -> 61 | Improved |
| share feature maps between detection and recognition (RPnet vs SSD300 + Holistic-CNN) | RPnet · CCPD recognition | AP (%): 95.2 -> 95.5 | Improved |
| replace TE2E with RPnet | TE2E · CCPD detection | FPS: 3 -> 61 | Improved |
| replace SSD300 with RPnet for detection | SSD300 · CCPD detection (IoU>0.7) | AP (%): 94.4 -> 94.5 | No clear effect |
| use joint optimization of detection and recognition (RPnet/TE2E) rather than detection alone | Faster-RCNN · CCPD detection | AP (%): 92.9 -> 94.5 | Improved |
| pre-train detection module from scratch on CCPD rather than ImageNet | RPnet · CCPD | detection bounding box | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- YOLO9000 — YOLO9000: Better, Faster, Stronger (CVPR 2017) · better on 2 of 2 shared results
- Faster-RCNN — Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (TPAMI 2016) · 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
- LSV-LP: Large-Scale Video-Based License Plate Detection and Recognition (TPAMI 2022 · similarity 0.89)
- Towards Ship License Plate Recognition in the Wild: A Large Benchmark and Strong Baseline (AAAI 2025 · similarity 0.87)
- License Plate Detection and Recognition in Unconstrained Scenarios (ECCV 2018 · similarity 0.86)
- Towards Human-Level License Plate Recognition (ECCV 2018 · similarity 0.85)
- LP-Diff: Towards Improved Restoration of Real-World Degraded License Plate (CVPR 2025 · similarity 0.85)
- Efficient License Plate Recognition via Holistic Position Attention (AAAI 2021 · similarity 0.84)
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.
