Research map / Label-efficient and robust learning / Knowledge distillation
Active Learning for Lane Detection: A Knowledge Distillation Approach
Fengchao Peng, Chao Wang, Jianzhuang Liu, Zhen Yang
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 reverse-nearest-neighbor diversity to KD uncertainty (Ours vs KD+KM) (ablation) | KD-only selection · CULane + PointLaneNet | F1-Score | Improved |
| add K-Means++ diversity to KD-only selection (ablation) | KD-only selection · CULane + PointLaneNet | F1-Score | Improved |
| use KD uncertainty alone instead of ensemble prediction gap (Ens) (ablation) | Active learning selection · CULane + PointLaneNet | F1-Score | Improved |
| combine uncertainty and diversity (Ours) versus uncertainty-only methods | Active learning selection · CULane + PointLaneNet | F1-Score | Improved |
| use our uncertainty with noise penalty instead of cross entropy | Entropy (Ent) · CULane first iteration | average lanes per selected image: 1.5 -> 3.1 | Improved |
| select images with the most lanes instead of random | Lane-count selection · CULane | F1-Score | Hurt |
| use loss-prediction (LLoss) instead of random selection | LLoss · CULane | F1-Score | Hurt |
| use ACD spatial entropy instead of plain entropy for lane detection | ACD · CULane | F1-Score | No clear effect |
| apply method to UFLD instead of PointLaneNet | Ours · CULane and LLAMAS + UFLD | F1-Score | Improved |
| apply method to 2D object detection instead of ACD | Ours · BDD100K + Faster-RCNN | mAP | Improved |
Closest papers by meaning
- LaneSegNet: Map Learning with Lane Segment Perception for Autonomous Driving (ICLR 2024 · similarity 0.89)
- Sketch and Refine: Towards Fast and Accurate Lane Detection (AAAI 2024 · similarity 0.88)
- GEONet: Global Enhancement and Optimization Network for Lane Detection (AAAI 2025 · similarity 0.88)
- Lane2Seq: Towards Unified Lane Detection via Sequence Generation (CVPR 2024 · similarity 0.88)
- LaneCPP: Continuous 3D Lane Detection using Physical Priors (CVPR 2024 · similarity 0.88)
- Selective Focus: Investigating Semantics Sensitivity in Post-training Quantization for Lane Detection (AAAI 2024 · 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.
