Abstract
Deep learning enables high accuracy in object detection in comparison with alternative methods. However deep learning based algorithms are often computationally expensive. That limits the use in many real world scenarios. For decades, researchers have been working on speeding up object detection. One bottleneck in current state-of-the-art methods is the region proposal generation stage as hundreds and thousands of proposed regions need to be processed before detection. Most of the regions are background areas which do not contribute to the actual detection. To improve the efficiency, we propose a region proposal network that can significantly reduce background while maintaining high accuracy. The comparison with SOTA methods shows that our network can be up to 70 times faster, since it only contains 1/15 to 1/150 parameters relative to these methods. The class IoU for MS COCO subsets achieves 40% to 70% and the inference speed on GTX 1080Ti can achieve above 1000 FPS performance. In addition, our study shows that high resolution input is not a must for high accuracy. The use of down-sampled images can further reduce computation costs while retaining or even improving accuracy.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Duan, L., Gu, J., Yang, Z., Miao, J., Ma, W., Wu, C.: Bio-inspired visual attention model and saliency guided object segmentation. In: Pan, J.-S., Krömer, P., Snášel, V. (eds.) Genetic and Evolutionary Computing, pp. 291–298. Springer International Publishing, Cham (2014). https://doi.org/10.1007/978-3-319-01796-9_31
Everingham, M., Van Gool, L., Williams, C., Winn, J., Zisserman, A.: The pascal visual object classes challenge 2012 (voc2012) results (2012). http://www.pascal-network.org/challenges/VOC/voc2011/workshop/index.html
Fattal, A.K., Karg, M., Scharfenberger, C., Adamy, J.: Saliency-guided region proposal network for CNN based object detection. In: 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC), pp. 1–8. IEEE (2017)
Feng, J., Wei, Y., Tao, L., Zhang, C., Sun, J.: Salient object detection by composition. In: 2011 International Conference on Computer Vision, pp. 1028–1035. IEEE (2011)
Girshick, R., Donahue, J., Darrell, T., Malik, J.: Rich feature hierarchies for accurate object detection and semantic segmentation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 580–587 (2014)
Guo, M., Zhao, Y., Zhang, C., Chen, Z.: Fast object detection based on selective visual attention. Neurocomputing 144, 184–197 (2014)
Hosang, J., Benenson, R., Dollár, P., Schiele, B.: What makes for effective detection proposals? IEEE Trans. Pattern Anal. Mach. Intell. 38(4), 814–830 (2015)
Hu, J.Y., Shi, C.J.R., Zhang, J.S.: Saliency-based yolo for single target detection. Knowl. Inf. Syst. 63(3), 717–732 (2021)
Jiang, Z., Zhao, L., Li, S., Jia, Y.: Real-time object detection method based on improved yolov4-tiny. arXiv preprint arXiv:2011.04244 (2020)
Lin, T.Y., Goyal, P., Girshick, R., He, K., Dollár, P.: Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2980–2988 (2017)
Lin, T.-Y., et al.: Microsoft COCO: common objects in context. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8693, pp. 740–755. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-10602-1_48
Ling, H., Qin, Y., Zhang, L., Shi, Y., Li, P.: Selective convolutional network: an efficient object detector with ignoring background. In: ICASSP 2020–2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 4462–4466. IEEE (2020)
Liu, W., et al.: SSD: single shot MultiBox detector. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) ECCV 2016. LNCS, vol. 9905, pp. 21–37. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46448-0_2
Long, X., et al.: PP-YOLO: an effective and efficient implementation of object detector. arXiv preprint arXiv:2007.12099 (2020)
Redmon, J., Divvala, S., Girshick, R., Farhadi, A.: You only look once: unified, real-time object detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 779–788 (2016)
Redmon, J., Farhadi, A.: YOLOv3: an incremental improvement. arXiv preprint arXiv:1804.02767 (2018)
Ren, S., He, K., Girshick, R., Sun, J.: Faster R-CNN: towards real-time object detection with region proposal networks. arXiv preprint arXiv:1506.01497 (2015)
Ren, Z., Gao, S., Chia, L.T., Tsang, I.W.H.: Region-based saliency detection and its application in object recognition. IEEE Trans. Circuits Syst. Video Technol. 24(5), 769–779 (2013)
Uijlings, J.R., Van De Sande, K.E., Gevers, T., Smeulders, A.W.: Selective search for object recognition. Int. J. Comput. Vision 104(2), 154–171 (2013)
Wang, W., Shen, J., Shao, L.: Video salient object detection via fully convolutional networks. IEEE Trans. Image Process. 27(1), 38–49 (2017)
Yohanandan, S., Song, A., Dyer, A.G., Tao, D.: Saliency preservation in low-resolution grayscale images. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds.) ECCV 2018. LNCS, vol. 11210, pp. 237–254. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01231-1_15
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 Springer Nature Switzerland AG
About this paper
Cite this paper
Li, W., Song, A. (2022). UFO RPN: A Region Proposal Network for Ultra Fast Object Detection. In: Long, G., Yu, X., Wang, S. (eds) AI 2021: Advances in Artificial Intelligence. AI 2022. Lecture Notes in Computer Science(), vol 13151. Springer, Cham. https://doi.org/10.1007/978-3-030-97546-3_50
Download citation
DOI: https://doi.org/10.1007/978-3-030-97546-3_50
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-97545-6
Online ISBN: 978-3-030-97546-3
eBook Packages: Computer ScienceComputer Science (R0)