Research map / Detection and segmentation / CNN features
Domain-specific segmentation and detection
Adapt context-aware networks to radar, maps, HDR, and medical imaging.
118 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in CNN features, keyword cluster “local · context · semantic segmentation”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
29 papers in 2023–24 and 44 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 8.8% to 12.7%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Mitosis Detection in Breast Cancer Histology Images via Deep Cascaded Networks (AAAI 2016 · 262 citations)
- 2017 · Deep Joint Rain Detection and Removal From a Single Image (CVPR 2017 · 1,280 citations)
- 2018 · DecideNet: Counting Varying Density Crowds Through Attention Guided Detection and Density Estimation (CVPR 2018 · 454 citations)
- 2019 · From Coarse to Fine: Robust Hierarchical Localization at Large Scale (CVPR 2019 · 1,054 citations)
- 2020 · Multi-Scale Progressive Fusion Network for Single Image Deraining (CVPR 2020 · 788 citations)
- 2021 · MP3: A Unified Model To Map, Perceive, Predict and Plan (CVPR 2021 · 223 citations)
- 2022 · ISNet: Shape Matters for Infrared Small Target Detection (CVPR 2022 · 715 citations)
- 2023 · SARAS-Net: Scale and Relation Aware Siamese Network for Change Detection (AAAI 2023 · 77 citations)
- 2024 · Modality-agnostic Domain Generalizable Medical Image Segmentation by Multi-Frequency in Multi-Scale Attention (CVPR 2024 · 107 citations)
- 2025 · HS-FPN: High Frequency and Spatial Perception FPN for Tiny Object Detection (AAAI 2025 · 108 citations)
- 2026 · Seeing Beyond Illusion: Generalized and Efficient Mirror Detection (AAAI 2026 · 4 citations)
Earliest papers here
- Mitosis Detection in Breast Cancer Histology Images via Deep Cascaded Networks (AAAI 2016 · 262 citations)
- Recombinator Networks: Learning Coarse-To-Fine Feature Aggregation (CVPR 2016 · 124 citations)
- Semantic Filtering (CVPR 2016 · 37 citations)
What has been tried
90 papers here report 947 measured changes (a component added, removed or swapped, compared with a baseline): helped 79%, no clear effect 6%, hurt 14%, mixed 1%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Add channel path only | AI-TOD test · AP | 20.2 -> 21.6 | Helped | HS-FPN: High Frequency and Spatial Perception FPN for Tiny Object Detection (AAAI 2025 · 108 citations) |
| combine multi-frequency and multi-scale features rather than using single-scale or single-frequency variants | CVC-ClinicDB and Kvasir · DSC | 90.7 (0.4), 92.3 (0.5), 92.3 (0.3) -> 93.9 (0.6), 90.7 (0.8) | Helped | Modality-agnostic Domain Generalizable Medical Image Segmentation by Multi-Frequency in Multi-Scale Attention (CVPR 2024 · 107 citations) |
| Add CR to Stage 1 Net_0 training | Kvasir-SEG · mDSC | 85.1 -> 86.1 | Helped | ConDSeg: A General Medical Image Segmentation Framework via Contrast-Driven Feature Enhancement (AAAI 2025 · 78 citations) |
| replace pixel-wise prediction with C3P | MAS3K · mIoU | 0.772 -> 0.789 | Helped | Fantastic Animals and Where to Find Them: Segment Any Marine Animal with Dual SAM (CVPR 2024 · 55 citations) |
| Add learnable intensity weights instead of fixing wa and wb to 1/2 | FMB · SCD | 1.58 -> 1.76 | Helped | Task-driven Image Fusion with Learnable Fusion Loss (CVPR 2025 · 46 citations) |
| add TCIA to the basic pyramid ViT baseline | IRSTD-1k · IoU | 62.82 -> 67.26 | Helped | TCI-Former: Thermal Conduction-Inspired Transformer for Infrared Small Target Detection (AAAI 2024 · 39 citations) |
| Add inter-layer consistency loss | GLH-water test set · IoU (%) | +5.89% | Helped | GLH-Water: A Large-Scale Dataset for Global Surface Water Detection in Large-Size Very-High-Resolution Satellite Imagery (AAAI 2024 · 38 citations) |
| Add scale pruning to Mesorch | Four-dataset average · F1 | 0.6771 -> 0.6762 | No clear effect | Mesoscopic Insights: Orchestrating Multi-Scale & Hybrid Architecture for Image Manipulation Localization (AAAI 2025 · 32 citations) |
| Add TDP loss | PASCAL VOC2012 x8 · mIoU | 53.3 -> 55.0 | Helped | Beyond Image Super-Resolution for Image Recognition with Task-Driven Perceptual Loss (CVPR 2024 · 24 citations) |
| Add Inner Attention to the baseline U2-Net encoder | MoNuSeg · IoU | 72.03 -> 72.81 | Helped | Rethinking Decoder Design: Improving Biomarker Segmentation Using Depth-to-Space Restoration and Residual Linear Attention (CVPR 2025 · 21 citations) |
| use local rather than global denoising | ISTD test, shadow regions · RMSE | 10.43 -> 9.40 | Helped | Recasting Regional Lighting for Shadow Removal (AAAI 2024 · 20 citations) |
| Replace the Wu et al. backbone with the SalM² backbone without semantic guidance | TrafficGaze · NSS | 5.82 -> 5.87 | Helped | SalM²: An Extremely Lightweight Saliency Mamba Model for Real-Time Cognitive Awareness of Driver Attention (AAAI 2025 · 19 citations) |
What reviewers will expect
Datasets and baselines named in the 24 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- nuScenes · 3 of 24 papers
- ISIC-2017 · 2 of 24 papers
- ISIC2018 · 2 of 24 papers
- Argoverse2 · 2 of 24 papers
Baselines
- U-Net · 5 of 24 papers
- SwinUNet · 3 of 24 papers
- MapTRv2 · 3 of 24 papers
- StreamMapNet · 3 of 24 papers
- UNETR++ · 2 of 24 papers
- H2Former · 2 of 24 papers
- SwinUNETR · 2 of 24 papers
- VmUnet · 2 of 24 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 73 papers; from OpenAlex and the papers' affiliations).
- Harbin Institute of Technology · 5 papers
- Chinese Academy of Sciences · 4 papers
- City University of Hong Kong · 4 papers
- Tsinghua University · 4 papers
- Peking University · 3 papers
- Shenzhen University · 3 papers
- Huazhong University of Science and Technology · 3 papers
- Sichuan University · 3 papers
- University of Chinese Academy of Sciences · 3 papers
- Zhejiang University · 2 papers
Use it
Download BibTeX of 38 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Context modeling for semantic segmentation (241 papers)
- Local features and context in CNNs (170 papers)
- Context for hard detection conditions (133 papers)
- Forgery detection (46 papers)
- Lane detection (32 papers)
Elsewhere in the topic
- Region-based and query-based detectors (183 papers)
- Convolution-transformer backbones (101 papers)
- Box regression and loss design (94 papers)
- Efficient DETR query design (85 papers)
- Boundary-aware instance segmentation (83 papers)
- Plain and high-resolution ViTs (69 papers)
- Arbitrary-shape scene text detection (61 papers)
- RGB salient object detection (61 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).