Research map / Detection and segmentation / Open-set and few-shot detection
Perceptual Generative Adversarial Networks for Small Object Detection
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use Perceptual GAN instead of Zhu et al. | traffic sign detection · Tsinghua-Tencent 100K small traffic signs | recall: 87 -> 89 | Helped |
| use Perceptual GAN instead of Zhu et al. | traffic sign detection · Tsinghua-Tencent 100K small traffic signs | accuracy: 82 -> 84 | Helped |
| use Perceptual GAN instead of Fast R-CNN | small traffic sign detection · VOC 2007 test boat | AP: 59.4% -> 69.4% | Helped |
| use Perceptual GAN instead of Fast R-CNN | small traffic sign detection · VOC 2007 test bottle | AP: 38.3% -> 60.2% | Helped |
| use Perceptual GAN instead of Fast R-CNN | small traffic sign detection · VOC 2007 test chair | AP: 42.8% -> 57.9% | Helped |
| use Perceptual GAN instead of Fast R-CNN | small traffic sign detection · VOC 2007 test plant | AP: 31.8% -> 41.8% | Helped |
| replace generated features with skip pooling (ablation) | Perceptual GAN · Tsinghua-Tencent 100K small traffic signs | recall: 89 -> 76 | Hurt |
| replace generated features with skip pooling (ablation) | Perceptual GAN · Tsinghua-Tencent 100K small traffic signs | accuracy: 84 -> 82 | Hurt |
| increase image resolution to 2048×2048 (ablation) | small-object detection pipeline · Tsinghua-Tencent 100K small traffic signs | accuracy: 84 -> 81 | Hurt |
| use multi-scale input images (ablation) | small-object detection pipeline · Tsinghua-Tencent 100K small traffic signs | accuracy: 84 -> 77 | Hurt |
| train generator and discriminator with alternative optimization (ablation) | Perceptual GAN · Tsinghua-Tencent 100K small traffic signs | recall: 80 -> 89 | Helped |
| train generator and discriminator with alternative optimization (ablation) | Perceptual GAN · Tsinghua-Tencent 100K small traffic signs | accuracy: 80 -> 84 | Helped |
Closest papers by meaning
- Finding Tiny Faces in the Wild With Generative Adversarial Network (CVPR 2018 · similarity 0.88)
- Small-GAN: Speeding up GAN Training using Core-Sets (ICML 2020 · similarity 0.87)
- Spatially Invariant Unsupervised Object Detection with Convolutional Neural Networks (AAAI 2019 · similarity 0.87)
- Generative Region-Language Pretraining for Open-Ended Object Detection (CVPR 2024 · similarity 0.84)
- SET: Spectral Enhancement for Tiny Object Detection (CVPR 2025 · similarity 0.84)
- Believing is Seeing: Unobserved Object Detection using Generative Models (CVPR 2025 · similarity 0.84)
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.