Research map / Label-efficient and robust learning / Domain adaptation and generalization
Master of All: Simultaneous Generalization of Urban-Scene Segmentation to All Adverse Weather Conditions
Nikhil Reddy, Abhinav Singhal, Abhishek Kumar, Mahsa Baktashmotlagh, Chetan Arora
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 class imbalance re-weighting to the loss (ablation) | MALL-domain (DeepLabv3+ MobileNet) · Nighttime Driving test | mIoU: 37.4 -> 38.4 | Improved |
| add sample importance weighting to the loss (ablation) | MALL-domain (DeepLabv3+ MobileNet) · Nighttime Driving test | mIoU: 36.1 -> 38.4 | Improved |
| add class imbalance re-weighting to the loss (ablation) | MALL-domain (DeepLabv3+ MobileNet) · Dark Zurich val | mIoU: 20.6 -> 21.7 | Improved |
| add sample importance weighting to the loss (ablation) | MALL-domain (DeepLabv3+ MobileNet) · Dark Zurich val | mIoU: 19.2 -> 21.7 | Improved |
| use MALL-sample instead of TENT | DeepLabv3+ MobileNet · Nighttime Driving test | mIoU: 33.3 -> 36.8 | Improved |
| use per-domain adaptation (MALL-domain) instead of per-sample | DeepLabv3+ MobileNet · Nighttime Driving test | mIoU: 36.8 -> 38.4 | Improved |
| adapt with MALL-domain on nighttime driving | DeepLabv3+ ResNet101 · Cityscapes day | mIoU: 78.5 -> 75.7 | Hurt |
| adapt with MALL-domain on nighttime driving | DeepLabv3+ MobileNet · Cityscapes day | mIoU: 61.6 -> 59.6 | Hurt |
| apply MALL-domain at test time | MobileNetV2 · Average adverse-condition mIoU | mIoU: 24.2 -> 38.6 | Improved |
| apply MALL-domain at test time | SegFormer-B0 · Average adverse-condition mIoU | mIoU: 37.3 -> 38.7 | Improved |
| apply MALL-domain at test time | RobustNet-ResNet101 · Average adverse-condition mIoU | mIoU: 38.8 -> 42.0 | Improved |
| apply MALL-domain at test time | IBNNet · Average adverse-condition mIoU | mIoU: 39.6 -> 41.3 | 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.
- + TENT — Tent: Fully Test-Time Adaptation by Entropy Minimization (ICLR 2021) · better on 6 of 6 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
- UniMix: Towards Domain Adaptive and Generalizable LiDAR Semantic Segmentation in Adverse Weather (CVPR 2024 · similarity 0.91)
- Semantic Segmentation in Multiple Adverse Weather Conditions with Domain Knowledge Retention (AAAI 2024 · similarity 0.90)
- Parsing All Adverse Scenes: Severity-Aware Semantic Segmentation with Mask-Enhanced Cross-Domain Consistency (AAAI 2024 · similarity 0.90)
- End-to-End Video Semantic Segmentation in Adverse Weather using Fusion Blocks and Temporal-Spatial Teacher-Student Learning (NeurIPS 2024 · similarity 0.88)
- No Thing, Nothing: Highlighting Safety-Critical Classes for Robust LiDAR Semantic Segmentation in Adverse Weather (CVPR 2025 · similarity 0.87)
- Learning Generalized Segmentation for Foggy-Scenes by Bi-directional Wavelet Guidance (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.
