Object Detection Algorithm with Dynamic Loss and Enhanced Feature Fusion

doi:10.3778/j.issn.1673-9418.2301006

Abstract

Abstract: Object detection is one of the hottest directions in the field of computer vision. In order to further improve the performance of the object detection algorithm, a dynamic intersection over union loss (DYIoU Loss) based on the intersection over union (IoU) is proposed to solve the limitations of the position loss function in the training process. The relationship between the internal components of the position loss function is fully considered, and the weight of the position loss components can be given dynamically at different stages of the training to more specifically constrain the network. This enables the network to optimize different parts more effectively during the early, middle, and late stages of training to better align with the characteristics of the object detection task. In addition, in order to solve the deficiency of the feature fusion stage in the object detection network, deformable convolution is applied to the PAN (path aggregation network) structure, and a deformable path aggregation network neck (DePAN Neck) that can be plugged in is designed to improve the model’s ability to fuse multi-scale features and improve its detection performance on small objects. The above methods are applied to YOLOv6 models of YOLOv6-N, YOLOv6-T and YOLOv6-S sizes, and rich experiments are designed on the COCO2017 dataset to validate the effectiveness. The results show an average increase of 2.0 percentage points in the average precision (mAP).

Key words: object detection, deep learning, position loss, feature fusion

摘要： 目标检测是计算机视觉领域最火热的方向之一，为了进一步提升目标检测算法性能，解决目标检测算法在训练过程中位置损失函数存在的局限性，提出了基于交并比（IoU）的动态交并比损失函数（DYIoU Loss），充分考虑到了位置损失函数内部各组成部分之间的联系，可以在训练的不同阶段，动态给予位置损失组成部分不同的权重，以达到更有针对性地约束网络的目的，即让网络在训练的初期、中期和后期能够更符合目标检测任务的特性去优化不同部分。其次，为了解决目标检测网络在特征融合环节存在的不足，将可变形卷积应用到PAN结构当中，设计了一种可以即插即用的基于可变形卷积的DePAN Neck来提升模型对多尺度特征的融合能力，进而提升模型对小目标物体的检测效果。将上述方法应用到YOLOv6-N、YOLOv6-T、YOLOv6-S三个体量的YOLOv6模型上，在COCO2017数据集上设计了丰富实验，验证方法的有效性，平均精度（mAP）平均提升2.0个百分点。

关键词: 目标检测, 深度学习, 位置损失, 特征融合

ZHAO Qiming, ZHANG Tao, SUN Jun. Object Detection Algorithm with Dynamic Loss and Enhanced Feature Fusion[J]. Journal of Frontiers of Computer Science and Technology, 2023, 17(12): 2942-2953.

肇启明, 张涛, 孙俊. 动态损失与加强特征融合的目标检测算法[J]. 计算机科学与探索, 2023, 17(12): 2942-2953.

References

[1] GIRSHICK R. Fast R-CNN[C]//Proceedings of the 2015 IEEE International Conference on Computer Vision, Santiago, Dec 7-13, 2015. Washington: IEEE Computer Society, 2015: 1440-1448.
[2] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unfied, real-time object detection[C] //Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, Jun 27-30, 2016. Washington: IEEE Computer Society, 2016: 779-788.
[3] REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, Jul 21-26, 2017. Washington: IEEE Computer Society, 2017: 6517-6525.
[4] REDMON J, FARHADI A. YOLOv3: an incremental improve-ment[J]. arXiv:1804.02767, 2018.
[5] LIN T Y, DOLLáR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, Jul 21-26, 2017. Washington: IEEE Computer Society, 2017: 2117-2125.
[6] HE K, ZHANG X, REN S, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(9): 1904-1916.
[7] LIU S, QI L, QIN H, et al. Path aggregation network for instance segmentation[C]//Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, Jun 18-22, 2018. Washington: IEEE Computer Society, 2018: 8759-8768.
[8] DALAL N, TRIGGS B. Histograms of oriented gradients for human detection[C]//Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Diego, Jun 20-26, 2005. Washington: IEEE Computer Society, 2005: 886-893.
[9] LOWE D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision, 2004, 60(2): 91-110.
[10] 范丽丽, 赵宏伟, 赵浩宇,等. 基于深度卷积神经网络的目标检测研究综述[J]. 光学精密工程, 2020, 28(5):1 152-1164.
FAN L L, ZHAO H W, ZHAO H Y, et al. Survey of target detection based on deep convolutional neural networks[J]. Optics and Precision Engineering, 2020, 28(5): 1152-1164.
[11] 史彩娟, 张卫明, 陈厚儒, 等. 基于深度学习的显著性目标检测综述[J]. 计算机科学与探索, 2021, 15(2): 219-232.
SHI C J, ZHANG W M, CHEN H R, et al. Survey of salient object detection based on deep learning[J]. Journal of Frontiers of Computer Science and Technology, 2021, 15(2): 219-232.
[12] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the 2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, Jun 23-28, 2014. Washington: IEEE Computer Society, 2014: 580-587.
[13] REN S, HE K, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 39(6): 1137-1149.
[14] LI C, LI L, JIANG H, et al. YOLOv6: a single-stage object detection framework for industrial applications[J]. arXiv:2209.02976, 2022.
[15] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision, Venice, Oct 22-29, 2017. Washington: IEEE Computer Society, 2017: 2980-2988.
[16] YU J, JIANG Y, WANG Z, et al. Unitbox: an advanced object detection network[C]//Proceedings of the 24th ACM International Conference on Multimedia, Amsterdam, Oct 15-19, 2016. New York: ACM, 2016: 516-520.
[17] REZATOFIGHI H, TSOI N, GWAK J Y, et al. Generalized intersection over union: a metric and a loss for bounding box regression[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Long Beach, Jun 16-20, 2019. Piscataway: IEEE, 2019: 658-666.
[18] ZHENG Z, WANG P, LIU W, et al. Distance-IoU loss: faster and better learning for bounding box regression[C]//Proceedings of the 34th AAAI Conference on Artificial Intelligence, the 32nd Innovative Applications of Artificial Intelligence Conference, the 10th AAAI Symposium on Educational Advances in Artificial Intelligence, New York, Feb 7-12, 2020. Menlo Park: AAAI, 2020: 12993-13000.
[19] GEVORGYAN Z. SIoU Loss: more powerful learning for bounding box regression[J]. arXiv:2205.12740, 2022.
[20] BOCHKOVSKIY A, WANG C Y, LIAO H Y M. YOLOv4: optimal speed and accuracy of object detection[J]. arXiv:2004.10934, 2020.
[21] DAI J, QI H, XIONG Y, et al. Deformable convolutional networks[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision, Venice, Oct 22-29, 2017. Washington: IEEE Computer Society, 2017: 764-773.
[22] LIN T Y, MAIRE M, BELONGIE S, et al. Microsoft coco: common objects in context[C]//LNCS 8693：Proceedings of the 13th European Conference on Computer Vision, Zurich, Sep 5-12, 2014. Cham: Springer, 2014: 740-755.
[23] EVERINGHAM M, ESLAMI S, GOOL L, et al. The PASCAL visual object classes challenge: a retrospective[J]. International Journal of Computer Vision, 2014, 111(1): 98-136.
[24] GE Z, LIU S, WANG F, et al. YOLOX: exceeding YOLO series in 2021[J]. arXiv:2107.08430, 2021.
[25] XU S, WANG X, LV W, et al. PP-YOLOE: an evolved version of YOLO[J]. arXiv:2203.16250, 2022.
[26] WANG C Y, BOCHKOVSKIY A, LIAO H Y M. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[J]. arXiv:2207.02696, 2022.