注意力增强Salience-DETR的小目标识别算法

Small Object Detection Algorithm for Attention Enhanced Salience-DETR

摘要

针对DETR(detection transformer)系列模型在密集场景下易出现目标遗漏及在复杂背景中泛化能力不足的问题,提出一种注意力增强的Salience DETR模型.通过在Salience DETR模型中嵌入前景注意力机制与空间和通道协同注意力机制,提升模型对小目标特征的提取能力和复杂环境下的多目标识别能力.在数据集Common Objects in Context,Vision Meets Drone,Insects上的实验结果表明,该模型相较于Salience DETR模型平均精度分别提升1.0百分点、1.2百分点和2.2百分点.在自采集烟虫数据集上的实验结果表明,该模型对小尺寸目标识别评价指标提升2.0百分点,验证了模型的泛化能力.因此,该模型提升了复杂环境下的检测性能,并为小目标识别及特定领域应用提供了可行的技术方案.

Abstract

Aiming at the problems of target omission in dense scenes and insufficient generalization ability in complex backgrounds for the detection transformer (DETR) series models, we proposed an attention-enhanced Salience DETR model. By embedding a foreground attention mechanism as well as spatial and channel joint attention mechanisms into the Salience DETR model, the model’s ability to extract features of small targets and recognize multiple targets in complex environments was improved. The experimental results on the Common Objects in Context dataset, Vision Meets Drone dataset, and Insects dataset show that the mean average precision of this model is improved by 1.0 percentage point, 1.2 percentage points, and 2.2 percentage points respectively compared to the Salience Detection Transformer model. The experimental results on a self-collected corn borer dataset show that the model improves the evaluation metrics for small target recognition by 2.0 percentage points, verifying the model’s generalization capability. Therefore, this model enhances detection performance in complex environments and provides a feasible technical solution for small target recognition and specific domain applications.

关键词

小目标识别 / 注意力机制 / 复杂背景 / 烟虫识别

Key words

small object detection / attention mechanism / complex background / tobacco pests identification

参考文献

[1] Wei W, Cheng Y, He J F, et al. A Review of Small Object Detection Based on Deep Learning[J]. Neural Computing and Applications, 2024, 36(12): 6283-6303.

[2] Mahaur B, Mishra K K, Kumar A. An Improved Lightweight Small Object Detection Framework Applied to Real-Time Autonomous Driving[J]. Expert Systems with Applications, 2023, 234: 121036-1-121036-19.

[3] Yang J H, Xie M F, Hu C P, et al. Deep Learning for Detecting Cerebral Aneurysms with CT Angiography[J]. Radiology, 2021, 298: 155-163.

[4] Girshick R. Fast R-CNN[C]// Proceedings of the IEEE International Conference on Computer Vision. Piscataway, NJ: IEEE, 2015: 1440-1448.

[5] Redmon J, Divvala S, Girshick R, et al. You Only Look Once: Unified, Real-Time Object Detection[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2016: 779-788.

[6] 潘晓英, 贾凝心, 穆元震, 等. 小目标检测研究综述[J]. 中国图象图形学报, 2023, 28(9):2587-2615.

[7] (Pan X Y, Jia N X, Mu Y Z, et al. Survey of Small Object Detection[J]. Journal of Image and Graphics, 2023, 28(9): 2587-2615.)

[8] Zhu M J. Dynamic Feature Pyramid Networks for Object Detection[C]// Fifteenth International Conference on Signal Processing Systems (ICSPS 2023). [S. l.]: SPIE, 2024: 503-511.

[9] Chen S F, Sun P Z, Song Y B, et al. Diffusiondet: Diffusion Model for Object Detection[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway, NJ: IEEE, 2023: 19830-19843.

[10] Yang C H Y, Huang Z H, Wang N Y. QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small Object Detection[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2022: 13668-13677.

[11] Wang H, Gao P. Survey of Small Object Detection Methods Based on Deep Learning[C]// 2024 9th International Conference on Intelligent Informatics and Biomedical Sciences. Piscataway, NJ: IEEE, 2024: 221-224.

[12] Zhu X Z, Su W J, Lu L W, et al. Deformable DETR: Deformable Transformers for End-to-End Object Detection [EB/OL]. (2020-10-08)[2025-03-01]. https://arxiv.org/abs/2010.04159.

[13] Paz D, Zhang H Y, Christensen H I. Tridentnet: A Conditional Generative Model for Dynamic Trajectory Generation[C]// International Conference on Intelligent Autonomous Systems. Berlin: Springer International Publishing, 2021: 403-416.

[14] Tan M X, Pang R M, Le Q V. Efficientdet: Scalable and Efficient Object Detection[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2020: 10781-10790.

[15] Liu Z, Lin Y T, Cao Y, et al. Swin Transformer: Hierarchical Vision Transformer Using Shifted Windows[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway, NJ: IEEE, 2021: 10012-10022.

[16] Hou X Q, Liu M Q, Zhang S L, et al. Salience DETR: Enhancing Detection Transformer with Hierarchical Salience Filtering Refinement[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2024: 17574-17583.

[17] Lin T Y, Maire M, Belongie S, et al. Microsoft COCO: Common Objects in Context[C]// European Conference on Computer Vision. Berlin: Springer International Publishing, 2014: 740-755.

[18] 山显英, 张琳, 李泽慧. 深度学习驱动下的目标检测研究进展综述[J]. 计算机工程与应用, 2025, 61(1):24-41.

[19] (Shan X Y, Zhang L, Li Z H. Review of Research Progress in Object Detection Driven by Deep Learning[J]. Computer Engineering and Applications, 2025, 61(1): 24-41.)

[20] 董亚超, 包俊, 刘宏哲. 基于深度学习的小目标检测技术的研究进展[J]. 计算机科学, 2019, 46(10A):168-172.

[21] (Dong Y C, Bao J, Liu H Z. Resacrch Progress of Small Object Detection Technology Based on Deep Learning[J]. Computer Science, 2019, 46(10A): 168-172.)

[22] Tong K, Wu Y Q, Zhou F. Recent Advances in Small Object Detection Based on Deep Learning: A Review[J]. Image and Vision Computing, 2020, 97: 103910-1-103910-14.

[23] Meng D P, Chen X K, Fan Z J, et al. Conditional DETR for Fast Training Convergence[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway, NJ: IEEE, 2021: 3651-3660.

[24] Liu S L, Li F, Zhang H, et al. Dab-DETR: Dynamic Anchor Boxes Are Better Queries for DETR [EB/OL]. (2022-01-28)[2025-03-12]. https://arxiv.org/abs/2201.12329.

[25] Li F, Zhang H, Liu S L, et al. DN-DETR: Accelerate DETR Training by Introducing Query Denoising[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2022: 13619-13627.

[26] Chen Q, Chen X K, Wang J, et al. Group DETR: Fast DETR Training with Group-Wise One-to-Many Assignment[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway, NJ: IEEE, 2023: 6633-6642.

[27] Carion N, Massa F, Synnaeve G, et al. End-to-End Object Detection with Transformers[C]// European Conference on Computer Vision. Berlin: Springer International Publishing, 2020: 213-229.

[28] Yuan L, Hou Q B, Jiang Z H, et al. VOLO: Vision Outlooker for Visual Recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 45(5): 6575-6586.

[29] Si Y Z, Xu H Y, Zhu X Z, et al. SCSA: Exploring the Synergistic Effects between Spatial and Channel Attention[J]. Neurocomputing, 2025, 634: 129866-1-129866-12.

[30] Liu W, Anguelov D, Erhan D, et al. SSD: Single Shot Multibox Detector[C]// European Conference on Computer Vision. Berlin: Springer, 2016: 21-37.

[31] Du D W, Zhu P F, Wen L Y, et al. VisDrone-DET2019: The Vision Meets Drone Object Detection in Image Challenge Results[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops. Piscataway, NJ: IEEE, 2019: 213-226.

[32] Xu C, Wang J W, Yang W, et al. Detecting Tiny Objects in Aerial Images: A Normalized Wasserstein Distance and a New Benchmark[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 190: 79-93.

[33] Kinga D P, Adam J B. A Method for Stochastic Optimization [EB/OL]. (2014-12-22)[2025-03-15]. https://arxiv.org/abs/1412.6980.

[34] Zheng D H, Dong W H, Hu H L, et al. Less Is More: Focus Attention for Efficient DETR[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Piscataway, NJ: IEEE, 2023: 6674-6683.

[35] Jia D, Yuan Y H, He H D, et al. DETRs with Hybrid Matching[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2023: 19702-19712.

[36] Zang H, Li F, Liu S L, et al. DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection [EB/OL]. (2022-03-07)[2025-03-20]. https://arxiv.org/abs/2203.03605.

基金资助

云南省重大科技专项计划项目(202402AD080003)