Research map / Label-efficient and robust learning / Semi-supervised learning
R$$^{2}$$S100K: Road-Region Segmentation Dataset for Semi-supervised Autonomous Driving in the Wild
Muhammad Atif Butt, Hassan Mohamed Ali, Adnan Qayyum, Waqas Sultani, Ala Al‐Fuqaha, Junaid Qadir
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 EDS sampling in place of STDS for 3K labeled images (ablation) | Deeplab-v3+ ResNet-101 · R2S100K | mIoU: 56.14 -> 62.86 | Improved |
| add EDS with 2K pseudo labels (ablation) | Student (PSPNet) · R2S100K | mIoU: 59.87 -> 63.15 | Improved |
| add EDS with 32K pseudo labels (ablation) | Student (PSPNet) · R2S100K | mIoU: 62.33 -> 67.40 | Improved |
| add pseudo labels (3K real + 32K pseudo) over teacher, with EDS (ablation) | Student · R2S100K | mIoU: 62.86 -> 67.40 | Improved |
| increase labeled set from 3K to 9K (ablation) | Deeplab-v3+ · R2S100K | mIoU: 58.02 -> 55.37 | Hurt |
| increase pseudo labels from 8K to 32K without EDS (ablation) | Student · R2S100K | mIoU: 63.50 -> 62.33 | No clear effect |
| add EDS to PseudoSeg (ablation) | PseudoSeg · R2S100K | mIoU: 64.89 -> 66.73 | Improved |
| add EDS to CCT (ablation) | CCT · R2S100K | mIoU: 65.14 -> 65.40 | No clear effect |
| add EDS to HLCon (ablation) | HLCon · R2S100K | mIoU: 66.38 -> 66.91 | Improved |
| add EDS to PSPNet student (ablation) | PSPNet · R2S100K test | mIoU: 64.51 -> 67.23 | Improved |
| add EDS to LRASPP student (ablation) | LRASPP · R2S100K test | mIoU: 59.87 -> 61.38 | Improved |
| add EDS to BiSeNet student (ablation) | BiSeNet · R2S100K test | mIoU: 63.17 -> 64.93 | Improved |
Closest papers by meaning
- A Dataset for Lane Instance Segmentation in Urban Environments (ECCV 2018 · similarity 0.89)
- Weakly Supervised Semantic Segmentation for Driving Scenes (AAAI 2024 · similarity 0.88)
- In Defense of Pre-Trained ImageNet Architectures for Real-Time Semantic Segmentation of Road-Driving Images (CVPR 2019 · similarity 0.88)
- LaneSegNet: Map Learning with Lane Segment Perception for Autonomous Driving (ICLR 2024 · similarity 0.87)
- Traffic Scene Parsing through the TSP6K Dataset (CVPR 2024 · similarity 0.87)
- Unveiling the Hidden: Online Vectorized HD Map Construction with Clip-Level Token Interaction and Propagation (NeurIPS 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.
