도로 노면 파손 영상의 다중 분류 심층 신경망 평가를 통한 Backbone Network 선정 기법
A Selection Method of Backbone Network through Multi-Classification Deep Neural Network Evaluation of Road Surface Damage Images
0
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
최근 들어 인공 지능을 이용한 영상 객체 인식에 대한 연구 및 개발이 활발하게 진행되고있다. 그 연장선상에서 도로 유지 및 관리 분야에도 관련 연구의 활용도가 크게 향상될 것으로기대된다. 그 중에서도 특히 도로 노면 파손 객체 인식 (Object Detection) 을 위한 인공 지능모델이 지속적으로 개발되고 있다. 이러한 객체 인식 알고리즘을 개발하려면 우선적으로 특징지도를 생성하는 Backbone Network가 반드시 필요한데, 본 논문에서는 이를 선정하는 방법을제안하고자 한다. 이를 위해 6,000여 장의 도로 노면 파손 영상 데이터를 확보하고, 근래에 많이 사용되는 4종류의 심층 신경망을 활용하여 성능을 비교한다. 3가지의 성능 평가 방법을 적용하여 심층 신경망의 특징을 분석하고 최적의 심층 신경망을 결정한다. 또한 하이퍼 파라미터의 최적 조율을 통해 성능을 향상시키고, 최종적으로 도로 노면 파손 영상 분류를 위하여85.9%의 정확도로 수행이 가능한 경량화된 Backbone Network용 심층 신경망을 제안한다.
Abstract
In recent years, research and development on image object recognition using artificial intelligence have been actively carried out, and it is expected to be used for road maintenance. Among them, artificial intelligence models for object detection of road surface are continuously introduced. In order to develop such object recognition algorithms, a backbone network that extracts feature maps is essential. In this paper, we will discuss how to select the appropriate neural network. To accomplish it, we compared with 4 different deep neural networks using 6,000 road surface damage images. Based on three evaluation methods for analyzing characteristics of neural networks, we propose a method to determine optimal neural networks. In addition, we improved the performance through optimal tuning of hyper-parameters, and finally developed a light backbone network that can achieve 85.9% accuracy of road surface damage classification.
참고문헌 (25)
- Badrinarayanan V, "SegNet : A deep convolutional encoder-decoder architecture for image segmentation", The IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(12), pp.2481~2495 (2017)
- 전찬준, "완전 합성곱 신경망을 활용한 자동 포트홀 탐지 기술의 개발 및 평가", 한국ITS학회 논문지, 17(5), pp.55~64 (2018), 한국ITS학회[DOI]
- Dozat T, "Incorporating nesterov momentum into adam", ~ (2016)
- Eisenbach M, "How to get pavement distress detection ready for deep learning? A systematic approach", Proc. International Joint Conference on Neural Networks (IJCNN), pp.2039~2047 (2017)
- Girshick R, "Fast r-cnn", Proc. The IEEE International Conference on Computer Vision (ICCV), pp.1440~1448 (2015)
- Glorot X, "Understanding the difficulty of training deep feedforward neural networks", Proc. the Thirteenth International Conference on Artificial Intelligence and Statistics, pp.249~256 (2010)
- He K, "Mask R-CNN", Proc. The IEEE International Conference on Computer Vision (ICCV), pp.2980~2988 (2017)
- He K, "Deep residual learning for image recognition", Proc. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.770~778 (2016)
- Huang G, "Densely connected convolutional networks", Proc. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.4700~4708 (2017)
- Jo Y, "Pothole detection system using black-box camera", Sensors, 15(11), pp.29316~29331 (2015)
- 조영태, "포트홀 탐지 정확도 향상을 위한 Saliency Map 기반 포트홀 탐지 알고리즘", 한국ITS학회 논문지, 15(4), pp.104~114 (2016), 한국ITS학회[DOI]
- Kim S. J, "A study of image processing and deep learning algorithm for efficient of mobile-based pothole detection system", Proc. Korea Software Congress 2018 (KSC), pp.1595~1597 (2018)
- Kingma D. P, "ADAM: a method for stochastic optimization", Proc. third International Conference on Learning Representations (ICLR), pp.1~15 (2015)
- Krizhevsky A, "Learning multiple layers of features from tiny images", Technical report, University of Toronto, 1(4), pp.7~66 (2009)
- Lee S. W, "A proposal of a real-time pothole detection system based on machine learning and computer vision", Proc. Korea Computer Congress 2018 (KCC), pp.7~9 (2018)
- Li Z, "Detnet: A backbone network for object detection", ~ (2018)
- Long J, "Fully convolutional networks for semantic segmentation", Proc. IEEE Conference on Computer Vision and Pattern Recognition, pp.3431~3440 (2015)
- Maeda H, "Road damage detection and classification using deep neural networks with smartphone images", Computer-Aided Civil Infrastructure Engineering, 33(12), pp.1127~1141 (2018)
- "Smart Construction Technology Road Map" (2018)[URL]
- Netzer Y, "Reading digits in natural images with unsupervised feature learning", NIPS workshop (2011)
- Ren S, "Faster R-CNN: towards real-time object detection with region proposal networks", Proc. Advances in Neural Information Processing Systems 28 (NIPS 2015), pp.91~99 (2015)
- Simonyan K, "Very deep convolutional networks for large-scale image recognition", ~ (2014)
- Yang X, "Automatic pixel-level crack detection and measurement using fully convolutional networks", Computer-Aided Civil and Infrastructure Engineering, 33(12), pp.1090~1109 (2018)
- Yu F, "Multi-scale context aggregation by dilated convolution", ~ (2015)
- Zhang A, "Automated pixel-level pavement crack detection on 3D asphalt surfaces using a deep-learning network", Computer-Aided Civil and Infrastructure Engineering, 32(10), pp.805~819 (2017)