Traffic Scene Classification on a Representation Budget

被引:15
作者
Sikiric, Ivan [1 ]
Brkic, Karla [2 ]
Bevandic, Petra [2 ]
Kreso, Ivan [2 ]
Krapac, Josip [3 ]
Segvic, Sinisa [2 ]
机构
[1] Mireo Dd, Zagreb 10000, Croatia
[2] Univ Zagreb, Fac Elect Engn & Comp, Zagreb 10000, Croatia
[3] Mobius Labs GmbH, D-10997 Berlin, Germany
关键词
Visualization; Training; Feature extraction; Image representation; Servers; Global Positioning System; Architecture; Computer vision; intelligent vehicles; image classification; FEATURES;
D O I
10.1109/TITS.2019.2891995
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Visual cues can be used alongside GPS positioning and digital maps to improve understanding of vehicle environment in fleet management systems. Such systems are limited both in terms of bandwidth and storage space, so minimizing the size of transmitted and stored visual data is a priority. In this paper, we present efficient strategies for computing very short image representations suitable for classifying various types of traffic scenes in fleet management systems. We anticipate that the set of interesting classes will change over time, so we consider image representations that can be trained without knowing the labels of the target dataset. We empirically evaluate and compare the presented methods on a contributed dataset of 11447 labeled traffic scenes. Our results indicate that excellent classification results can be achieved with very short image representations, and that fine-tuning on the target dataset image data is not mandatory. Image descriptors can be as short as 128 components while still offering good performance, even in presence of adverse weather or illumination conditions.
引用
收藏
页码:336 / 345
页数:10
相关论文
共 43 条
[1]  
[Anonymous], 2016, P IVS WORKSH
[2]  
[Anonymous], 2016, RESFEATS RESIDUAL NE
[3]  
[Anonymous], IEEE T INTELL TRANSP
[4]  
[Anonymous], 2010, INT J COMPUT VISION, DOI DOI 10.1007/s11263-009-0275-4
[5]  
[Anonymous], P CVPR WORKSH VPRCE
[6]  
[Anonymous], 2011, ACM T INTEL SYST TEC, DOI DOI 10.1145/1961189.1961199
[7]  
Arjovsky M., 2017, P 34 INT C MACH LEAR, V70, P214
[8]   The devil is in the details: an evaluation of recent feature encoding methods [J].
Chatfield, Ken ;
Lempitsky, Victor ;
Vedaldi, Andrea ;
Zisserman, Andrew .
PROCEEDINGS OF THE BRITISH MACHINE VISION CONFERENCE 2011, 2011,
[9]  
Christian S., 2015, P IEEE C COMP VIS PA, DOI [DOI 10.1109/CVPR.2015.7298594, 10.1109/CVPR.2015.7298594]
[10]   Deep Filter Banks for Texture Recognition, Description, and Segmentation [J].
Cimpoi, Mircea ;
Maji, Subhransu ;
Kokkinos, Iasonas ;
Vedaldi, Andrea .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 2016, 118 (01) :65-94