Author(s)

MingYu Gao

Affiliation(s)

School of Mathematics and Physics, Xi'an Jiaotong-Liverpool University, Suzhou 215123, Jiangsu, China.

Corresponding Author

MingYu Gao

ABSTRACT

As convolutional neural networks continue to advance, more excellent models for image recognition have emerged in the field of computer vision. These models can help doctors identify causes of diseases in the medical field, reduce accidents in the transportation field, and collect facial recognition information in the security field. This study mainly focuses on the improvements of YOLOv8 compared to previous versions and its performance after training on the COCO dataset. It also briefly discusses the comparative results of YOLO with RCNN and SSD. Additionally, the development history of YOLO is introduced, with an emphasis on the performance of YOLOv8 after training and analysis the data. In the end, we see the future prospects of YOLO algorithm. 

KEYWORDS

YOLO; COCO dataset; Images

CITE THIS PAPER

MingYu Gao. YOLO and COCO dataset detective porfermance. World Journal of Information Technology. 2024, 2(2): 28-34. DOI: 10.61784/wjit3002.

REFERENCES

[1] Girshick R. Fast R-CNN. Proceedings of the IEEE International Conference on Computer Vision, 2015, 1440-1448.

[2] Liu W,Anguelov D, Erhan D, et al. SSD: Single Shot MultiBox Detector. Proceedings of the European Conference on Computer Vision, 2016, 21-37.

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

[4] Redmon J, Divvala S, Girshick R, et al. You only look once: Unified, real-time object detection. arXiv preprint arXiv:1506.02640, 2015.

[5] Redmon J, Farhadi A, YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, 7263-7271.

[6] Redmon J, Farhadi A. YOLOv3: An Incremental Improvement. arXiv preprint arXiv:1804.02767, 2018.

[7] Bochkovskiy A, Wang CY, Liao HYM. YOLOv4: Optimal Speed and Accuracy of Object Detection. arXiv preprint arXiv:2004.10934, 2020.

[8] Jocher G, et al. YOLOv5: Open-Source Object Detection.GitHub repository. The open-source implementation of YOLOv5 provided a foundation for the development of YOLOv8, 2021.

[9] Lin TY, Maire M, Belongie S, et al. Microsoft coco: Common objects in context. in Computer Vision–ECCV 2014: 13th European Conference, 2014, 740–755.

[10] Wang CY, Bochkovskiy A, Liao HYM. Scaled-YOLOv4: Scaling Cross Stage Partial Network. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, 13029-13038.

[11] Jocher G, et al.YOLOv5: Open-Source Object Detection. GitHub repository.

[12] Ge Z, Liu S, Wang F, et al. YOLOX: Exceeding YOLO Series in 2021. arXiv preprint arXiv:2107.08430, 2021.

[13] Chen K, et al. Hybrid Task Cascade for Instance Segmentation. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2019, 2961-2970.

[14] Liu W, et al. SSD: Single Shot MultiBox Detector. European Conference on Computer Vision (ECCV), 2016, 21-37.

[15] Lin TY, et al. Focal Loss for Dense Object Detection. Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2017, 2980-2988.

[16] Zoph B, et al. Learning Transferable Architectures for Scalable Image Recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, 8697-8710.

[17] Ren S, He K, Girshick R, et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. Advances in Neural Information Processing Systems (NeurIPS), 2015, 28, 91-99.

[18] Kumar P, kumar V. Exploring the Frontier of Object Detection: A Deep Dive into YOLOv8 and the COCO Dataset. in Proceedings of the IEEE conference on computer vision and Machine Intelligence (CVMI), 2023.