Interactive Transformer for Small Object Detection
Source abstract
Abstract
The detection of large-scale objects has achieved high accuracy, but due to the low peak signal to noise ratio (PSNR), fewer distinguishing features, and ease of being occluded by the surroundings, the detection of small objects, however, does not enjoy similar success. Endeavor to solve the problem, this paper proposes an attention mechanism based on cross-Key values. Based on the traditional transformer, this paper first improves the feature processing with the convolution module, effectively maintaining the local semantic context in the middle layer, and significantly reducing the number of parameters of the model. Then, to enhance the effectiveness of the attention mask, two Key values are calculated simultaneously along Query and Value by using the method of dual-branch parallel processing, which is used to strengthen the attention acquisition mode and improve the coupling of key information. Finally, focusing on the feature maps of different channels, the multi-head attention mechanism is applied to the channel attention mask to improve the feature utilization effect of the middle layer. By comparing three small object datasets, the plug-and-play interactive transformer (IT-transformer) module designed by us effectively improves the detection results of the baseline.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Advanced Neural Network Applications · Visual Attention and Saliency Detection · Advanced Image and Video Retrieval Techniques
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (40)
Show source-provided references
1. DKTNet: Dual-key transformer network for small object detection10.1016/j.neucom.2023.01.055 ↗
2. Dynamic coarse-to-fine learning for oriented tiny object detection
3. Multi background island bird detection based on faster R-CNN10.1080/01969722.2020.1827799 ↗
4. AugFPN: Improving multi-scale feature learning for object detection
5. NASFPN: Learning scalable feature pyramid architecture for object detection
6. ScaleKD: Distilling scale-aware knowledge in small object detector
7. GraphFPN: Graph feature pyramid network for object detection
8. Hu · 2021 · 15343
9. QueryDet: Cascaded sparse query for accelerating high-resolution small object detection
10. Focal loss for dense object detection
11. J. Redmon and A. Farhadi, “YOLOv3: An incremental improvement,” arXiv preprint arXiv:1804.02767, 2018.
12. Z. Ge, S. Liu, F. Wang, Z. Li and J. Sun, “YOLOX: Exceeding yolo series in 2021,” arXiv preprint arXiv:2107.08430, 2021.
13. Deformable convolutional networks
14. Deformable ConvNets V2: More deformable better results
15. Adaptive graph convolutional network with attention graph clustering for co-saliency detection
16. Attentional pyramid pooling of salient visual residuals for place recognition
17. TPH-YOLOv5: Improved YOLOv5 based on transformer prediction head for object detection on drone-captured scenarios
18. Attention is all you need
19. A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai et al., “An image is worth 16 × 16 words: Transformers for image recognition at scale,” arXiv preprint arXiv:2010.11929, 2020.
20. Swin transformer: Hierarchical vision transformer using shifted windows
21. X. Zhu, W. Su, L. Lu, B. Li, X. Wang et al., “Deformable DETR: Deformable transformers for end-to-end object detection,” arXiv preprint arXiv:2010.04159, 2020.
22. A. Bochkovskiy, C. Wang and H. M. Liao, “YOLOv4: Optimal speed and accuracy of object detection,” arXiv preprint arXiv:2004.10934, 2020.
23. YOLO9000: Better faster stronger
24. SSD: Single shot multibox detector
25. Focal loss for dense object detection
26. Faster R-CNN: Towards real-time object detection with region proposal networks
27. Cascade R-CNN: High quality object detection and instance segmentation10.1109/TPAMI.2019.2956516 ↗
28. H. Zhang, F. Li, S. Liu, L. Zhang, H. Su et al., “DINO: Detr with improved denoising anchor boxes for end-to-end object detection,” arXiv Preprint arXiv:2203.03605, 2022.
29. S. Chen, P. Sun, Y. Song and P. Luo, “DiffusionDet: Diffusion model for object detection,” arXiv preprint arXiv:2211.09788, 2022.10.1109/ICCV51070.2023.01816 ↗
30. I. Ethem Hamamci, S. Er, E. Simsar, A. Sekuboyina, M. Gundogar et al., “Diffusion-based hierarchical multi-label object detection to analyze panoramic dental X-rays,” arXiv preprint arXiv: 2303.06500, 2023.10.1007/978-3-031-43987-2_38 ↗
31. S. Nag, X. Zhu, J. Deng, Y. Song and T. Xiang, “DiffTAD: Temporal action detection with proposal denoising diffusion,” arXiv preprint arXiv:2303.14863, 2023.10.1109/ICCV51070.2023.00951 ↗
32. A novel data augmentation scheme for pedestrian detection with attribute preserving GAN10.1016/j.neucom.2020.02.094 ↗
33. Edge YOLO: Real-time intelligent object detection system based on edge-cloud cooperation in autonomous vehicles10.1109/TITS.2022.3158253 ↗
34. A full data augmentation pipeline for small object detection based on generative adversarial networks10.1016/j.patcog.2022.108998 ↗
35. An improved algorithm for small object detection based on YOLO v4 and multi-scale contextual information10.1016/j.compeleceng.2022.108490 ↗
36. SCRDet: Towards more robust detection for small, cluttered and rotated objects
37. Attention-based domain adaptation for single-stage detectors10.1007/s00138-022-01320-y ↗
38. Automatic detection of hardhats worn by construction personnel: A deep learning approach and benchmark dataset10.1016/j.autcon.2019.102894 ↗
39. HRDNet: High-resolution detection network for small objects
40. Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection