Triplet Loss with Channel Attention for Person Re-identification

Triplet Loss with Channel Attention for Person Re-identification

Abstract

The triplet loss function has seen extensive use within person re-identification. Most works focus on either improving the mining algorithm or adding new terms to the loss function itself. Our work instead concentrates on two other core components of the triplet loss that have been under-researched. First, we improve the standard Euclidean distance with dynamic weights, which are selected based on the standard deviation of features across the batch. Second, we exploit channel attention via a squeeze and excitation unit in the backbone model to emphasise important features throughout all layers of the model. This ensures that the output feature vector is a better representation of the image, and is also more suitable to use within our dynamically weighted Euclidean distance function. We demonstrate that our alterations provide significant performance improvement across popular reidentification data sets, including almost 10% mAP improvement on the CUHK03 data set. The proposed model attains results competitive with many state-of-the-art person re-identification models.

Publication

Daniel Organisciak, Chirine Riachy, Nauman Aslam and Hubert P. H. Shum,
"Triplet Loss with Channel Attention for Person Re-identification",
Journal of WSCG - Proceedings of the 2019 International Conferences in Central Europe on Computer Graphics, Visualization and Computer Vision (WSCG)

Downloads

YouTube

References

BibTeX

@article{organisciak19triplet,
 author={Organisciak, Daniel and Riachy, Chirine and Aslam, Nauman and Shum, Hubert P. H.},
 journal={Journal of WSCG},
 title={Triplet Loss with Channel Attention for Person Re-identification},
 year={2019},
 volume={27},
 number={2},
 pages={161--169},
 numpages={9},
 issn={1213-6972},
 location={Plzen, Czech Republic},
}

EndNote/RefMan

TY  - JOUR
AU  - Organisciak, Daniel
AU  - Riachy, Chirine
AU  - Aslam, Nauman
AU  - Shum, Hubert P. H.
T2  - Journal of WSCG
TI  - Triplet Loss with Channel Attention for Person Re-identification
PY  - 2019
VL  - 27
IS  - 2
SP  - 161
EP  - 169
SN  - 1213-6972
ER  - 

Plain Text

Daniel Organisciak, Chirine Riachy, Nauman Aslam and Hubert P. H. Shum, "Triplet Loss with Channel Attention for Person Re-identification," Journal of WSCG, vol. 27, no. 2, pp. 161-169, Plzen, Czech Republic, 2019.

Similar Research

 

         

Last update: 15/10/2019