Deep Ensemble Effectively and Efficiently for Vehicle Instance Retrieval - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2018

Deep Ensemble Effectively and Efficiently for Vehicle Instance Retrieval

Zhengyan Ding
  • Function : Author
  • PersonId : 1037185
Xiaoteng Zhang
  • Function : Author
  • PersonId : 1037186
Shaoxi Xu
  • Function : Author
  • PersonId : 1037187
Lei Song
  • Function : Author
  • PersonId : 1037188
Na Duan
  • Function : Author
  • PersonId : 1037189


This paper aims to highlight instance retrieval tasks centered around ‘vehicle’, due to its wide range of applications in surveillance scenario. Recently, image representations based on the convolutional neural network (CNN) have achieved significant success for visual recognition, including instance retrieval. However, many previous retrieval methods have not exploit the ensemble abilities of different models, which achieve limited accuracy since a certain kind of visual representation is not comprehensive. So we propose a Deep Ensemble Efficiently and Effectively (DEEE) framework, to preserve the impressive performance of deep representations and combine various deep architectures in a complementary way. It is demonstrated that a large improvement can be acquired with slight increase on computation. Finally, we evaluate the performance on two public vehicle datasets, VehicleID and VeRi, both outperforming state-of-the-art methods by a large margin.
Fichier principal
Vignette du fichier
470174_1_En_17_Chapter.pdf (581.62 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01888642 , version 1 (05-10-2018)





Zhengyan Ding, Xiaoteng Zhang, Shaoxi Xu, Lei Song, Na Duan. Deep Ensemble Effectively and Efficiently for Vehicle Instance Retrieval. 11th International Conference on Research and Practical Issues of Enterprise Information Systems (CONFENIS), Oct 2017, Shanghai, China. pp.192-201, ⟨10.1007/978-3-319-94845-4_17⟩. ⟨hal-01888642⟩
70 View
37 Download



Gmail Facebook Twitter LinkedIn More