OBJECT DETECTION IN TRAFFIC SCENES WITH CONVOLUTIONAL NEURAL NETWORKS
DOI:
https://doi.org/10.24867/13BE41PantovicKeywords:
Bounding box, Classification, Convolutional neural network, Object detectionAbstract
Object detection is the key technology behind advanced driver assistance systems. This paper demonstrates an application of convolutional neural networks in the task of detecting objects of interest from the sequence of images. Labeled images with various driving scenarios recorded during daylight city drive are taken from CrowdAI database. Images were used for model training and evaluation during testing phase. Results obtained using models with different number of convolutional layers and various activation functions are analyzed with purpose of using these models for real-time object detection.
References
[1] L. Hunjung, „Awesome vehicle datasets“ (CrowdAI baza podataka), https://github.com/hunjung-lim/awesome-vehicle-datasets
[2] J. Brownlee, „A Gentle Introduction to Object Recognition With Deep Learning“, https://machinelearningmastery.com/object-recognition-with-deep-learning. (poslednji pristup u martu 2021. godine)
[3] Y. LeCun, L. Bottou, Y. Bengio, P. Haffner, “Gradient-based learning applied to document recognition”, Proceedings of the IEEE, 86(11), pp. 2278–2324, 1998.
[4] A. Zhang, Z.C. Lipton, M. Li, A.J. Sola, „Dive into Deep Learning“, https://d2l.ai/ (poslednji pristup u martu 2021. godine)
[5] J. Redmon, S. Divvala, R. Girshick, A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.
[6] F.N. Iandola, S. Han, M.W. Moskewicz, K. Ashraf, W.J. Dally, K. Keutzer, “SqueezeNet: AlexNet-Level Accuracy with 50x Fewer Parameters and <0.5MB Model Size”, arXiv Prepr. arXiv1602.07360, 2016.
[7] K. He, J. Sun, “Convolutional neural networks at constrained time cost“, Conference on Computer Vision and Pattern Recognition(CVPR), 2015.
[2] J. Brownlee, „A Gentle Introduction to Object Recognition With Deep Learning“, https://machinelearningmastery.com/object-recognition-with-deep-learning. (poslednji pristup u martu 2021. godine)
[3] Y. LeCun, L. Bottou, Y. Bengio, P. Haffner, “Gradient-based learning applied to document recognition”, Proceedings of the IEEE, 86(11), pp. 2278–2324, 1998.
[4] A. Zhang, Z.C. Lipton, M. Li, A.J. Sola, „Dive into Deep Learning“, https://d2l.ai/ (poslednji pristup u martu 2021. godine)
[5] J. Redmon, S. Divvala, R. Girshick, A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.
[6] F.N. Iandola, S. Han, M.W. Moskewicz, K. Ashraf, W.J. Dally, K. Keutzer, “SqueezeNet: AlexNet-Level Accuracy with 50x Fewer Parameters and <0.5MB Model Size”, arXiv Prepr. arXiv1602.07360, 2016.
[7] K. He, J. Sun, “Convolutional neural networks at constrained time cost“, Conference on Computer Vision and Pattern Recognition(CVPR), 2015.
Downloads
Published
2021-07-04
Issue
Section
Electrotechnical and Computer Engineering
