Research map / Detection and segmentation / CNN features
Multitask AET With Orthogonal Tangent Regularity for Dark Object Detection
Ziteng Cui, Guo-Jun Qi, Lin Gu, Shaodi You, Zenghui Zhang, Tatsuya Harada
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 orthogonal tangent regularity (ablation) | MAET (YOLOv3) · VOC07 test (synthetic low-light) | AP50: 0.770 -> 0.788 | Improved |
| add orthogonal tangent regularity (ablation) | MAET (YOLOv3) · COCO val (synthetic low-light) | AP: 0.321 -> 0.330 | Improved |
| add orthogonal tangent regularity (ablation) | MAET (YOLOv3) · COCO val (synthetic low-light) | AP50: 0.534 -> 0.569 | Improved |
| add orthogonal tangent regularity (ablation) | MAET (YOLOv3) · COCO val (synthetic low-light) | AP_S: 0.163 -> 0.189 | Improved |
| add orthogonal tangent regularity (ablation) | MAET (YOLOv3) · ExDark | mAP: 0.722 -> 0.740 | Improved |
| add orthogonal tangent regularity (ablation) | MAET (YOLOv3) · UG2+ DARK FACE | mAP: 0.542 -> 0.558 | Improved |
| add degradation-decoding task (MAET) without ort, versus detector trained on synthetic low-light only (ablation) | YOLOv3 · VOC07 test (synthetic low-light) | AP50: 0.764 -> 0.770 | Improved |
| add degradation-decoding task (MAET) without ort, versus detector trained on synthetic low-light only (ablation) | YOLOv3 · ExDark | mAP: 0.716 -> 0.722 | No clear effect |
| add degradation-decoding task (MAET) without ort, versus detector trained on synthetic low-light only (ablation) | YOLOv3 · UG2+ DARK FACE | mAP: 0.540 -> 0.542 | No clear effect |
| use our ISP-based synthesis instead of Retinex synthesis (ablation) | YOLO (L) · ExDark | mAP: 0.698 -> 0.716 | Improved |
| use our ISP-based synthesis instead of Retinex synthesis (ablation) | YOLO (L) · UG2+ DARK FACE | mAP: 0.511 -> 0.540 | Improved |
| use our synthesis instead of inverse gamma with mixed Gaussian-Poisson noise (ablation) | YOLO (L) · UG2+ DARK FACE | mAP: 0.535 -> 0.540 | Improved |
Closest papers by meaning
- ISP-Teacher:Image Signal Process with Disentanglement Regularization for Unsupervised Domain Adaptive Dark Object Detection (AAAI 2024 · similarity 0.85)
- Unleashing Vanilla Vision Transformer with Masked Image Modeling for Object Detection (ICCV 2023 · similarity 0.84)
- Trash to Treasure: Low-Light Object Detection via Decomposition-and-Aggregation (AAAI 2024 · similarity 0.82)
- Closest Neighbors are Harmful for Lightweight Masked Auto-encoders (CVPR 2025 · similarity 0.81)
- A New Multi-Source Light Detection Benchmark and Semi-Supervised Focal Light Detection (NeurIPS 2024 · similarity 0.81)
- FRBNet: Revisiting Low-Light Vision through Frequency-Domain Radial Basis Network (NeurIPS 2025 · similarity 0.81)
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.
