Research map / Detection and segmentation / CNN features
HumanLiker: A Human-like Object Detector to Model the Manual Labeling Process
Haoran Wei, Ping Guo, Yangguang Zhu, Chenglong Liu, Peng Wang
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 first-stage top-left offset adjustment (ablation) | HumanLiker · COCO val2017, Swin-T | AP: 46.7 -> 47.1 | Improved |
| Add stage-two fine trimming (regression) branch (ablation) | HumanLiker · COCO val2017, Swin-T | AP: 45.4 -> 47.1 | Improved |
| Use multiple cascade blocks instead of one (ablation) | HumanLiker · COCO val2017, Swin-T | AP: 46.2 -> 47.1 | Improved |
| Use multiple cascade blocks instead of one (ablation) | HumanLiker · COCO val2017, Swin-T | FPS: 20 -> 18 | Hurt |
| Add corner pooling (CP) (ablation) | HumanLiker · COCO val2017, Swin-T | AP: 47.1 -> 47.1 | No clear effect |
| Add corner pooling (CP) (ablation) | HumanLiker · COCO val2017, Swin-T | FPS: 18 -> 15 | Hurt |
| Add criss-cross attention (CCA) (ablation) | HumanLiker · COCO val2017, Swin-T | AP: 47.1 -> 47.0 | No clear effect |
| Add criss-cross attention (CCA) (ablation) | HumanLiker · COCO val2017, Swin-T | FPS: 18 -> 16 | Hurt |
| Estimate top-left corner instead of center as keypoint (ablation) | keypoint model · MS-COCO training loss | training loss | Improved |
| Replace center-guided Deformable DETR with HumanLiker (ResNeXt-101-DCN) | HumanLiker · COCO test-dev | AP: 50.1 -> 50.2 | Improved |
| Replace Deformable DETR with HumanLiker | HumanLiker · COCO test-dev | AP50: 69.7 -> 67.9 | Hurt |
| Replace Cascade R-CNN with HumanLiker | HumanLiker · COCO test-dev | AP75: 52.9 -> 54.8 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- Deformable DETR — Deformable DETR: Deformable Transformers for End-to-End Object Detection (ICLR 2021) · better on 3 of 3 shared results
- Cascade R-CNN — Cascade R-CNN: Delving Into High Quality Object Detection (CVPR 2018) · better on 3 of 3 shared results
- Sparse R-CNN — Sparse R-CNN: End-to-End Object Detection With Learnable Proposals (CVPR 2021) · better on 1 of 1 shared result
- CentripetalNet — CentripetalNet: Pursuing High-Quality Keypoint Pairs for Object Detection (CVPR 2020) · better on 1 of 1 shared result
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
- Detecting and Recognizing Human-Object Interactions (CVPR 2018 · similarity 0.86)
- ObjectNet: A large-scale bias-controlled dataset for pushing the limits of object recognition models (NeurIPS 2019 · similarity 0.86)
- Detecting Human-Object Interactions via Functional Generalization (AAAI 2020 · similarity 0.86)
- LINK: Learning Instance-level Knowledge from Vision-Language Models for Human-Object Interaction Detection (ICLR 2026 · similarity 0.86)
- Roboflow100-VL: A Multi-Domain Object Detection Benchmark for Vision-Language Models (NeurIPS 2025 · similarity 0.85)
- Object Detection Benchmarks are Incomplete: The Role of Label Errors and Annotation Uncertainty (NeurIPS 2026 · similarity 0.85)
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.
