TY - GEN
T1 - Traffic scene recognition based on deep CNN and VLAD spatial pyramids
AU - Wu, Fang Yu
AU - Yan, Shi Yang
AU - Smith, Jeremy S.
AU - Zhang, Bai Ling
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2017/11/14
Y1 - 2017/11/14
N2 - Traffic scene recognition is an important and challenging issue in Intelligent Transportation Systems (ITS). Recently, Convolutional Neural Network (CNN) models have achieved great success in many applications, including scene classification. The remarkable representational learning capability of CNN remains to be further explored for solving real-world problems. Vector of Locally Aggregated Descriptors (VLAD) encoding has also proved to be a powerful method in catching global contextual information. In this paper, we attempted to solve the traffic scene recognition problem by combining the features representational capabilities of CNN with the VLAD encoding scheme. More specifically, the CNN features of image patches generated by a region proposal algorithm are encoded by applying VLAD, which subsequently represent an image in a compact representation. To catch the spatial information, spatial pyramids are exploited to encode CNN features. We experimented with a dataset of 10 categories of traffic scenes, with satisfactory categorization performances.
AB - Traffic scene recognition is an important and challenging issue in Intelligent Transportation Systems (ITS). Recently, Convolutional Neural Network (CNN) models have achieved great success in many applications, including scene classification. The remarkable representational learning capability of CNN remains to be further explored for solving real-world problems. Vector of Locally Aggregated Descriptors (VLAD) encoding has also proved to be a powerful method in catching global contextual information. In this paper, we attempted to solve the traffic scene recognition problem by combining the features representational capabilities of CNN with the VLAD encoding scheme. More specifically, the CNN features of image patches generated by a region proposal algorithm are encoded by applying VLAD, which subsequently represent an image in a compact representation. To catch the spatial information, spatial pyramids are exploited to encode CNN features. We experimented with a dataset of 10 categories of traffic scenes, with satisfactory categorization performances.
KW - Convolutional Neural Network
KW - Traffic scene recognition
KW - Vector of Locally Aggregated Descriptors encoding
UR - http://www.scopus.com/inward/record.url?scp=85042518576&partnerID=8YFLogxK
U2 - 10.1109/ICMLC.2017.8107758
DO - 10.1109/ICMLC.2017.8107758
M3 - Conference Proceeding
AN - SCOPUS:85042518576
T3 - Proceedings of 2017 International Conference on Machine Learning and Cybernetics, ICMLC 2017
SP - 156
EP - 161
BT - Proceedings of 2017 International Conference on Machine Learning and Cybernetics, ICMLC 2017
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 16th International Conference on Machine Learning and Cybernetics, ICMLC 2017
Y2 - 9 July 2017 through 12 July 2017
ER -