Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-Identification
This study aims to leverage Binary Neural Networks (BNN) to learn binary hash codes for efficient person re-identification (ReID). BNNs, which use binary weights and activations, show promise in speeding up the inference time in deep models. However, BNNs typically suffer from performance degradatio...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10138817/ |
_version_ | 1797799317139881984 |
---|---|
author | Ayse Serbetci Yusuf Sinan Akgul |
author_facet | Ayse Serbetci Yusuf Sinan Akgul |
author_sort | Ayse Serbetci |
collection | DOAJ |
description | This study aims to leverage Binary Neural Networks (BNN) to learn binary hash codes for efficient person re-identification (ReID). BNNs, which use binary weights and activations, show promise in speeding up the inference time in deep models. However, BNNs typically suffer from performance degradation mainly due to the discontinuity of the binarization operation. Proxy functions have been proposed to calculate the gradients in the backward propagation, but they lead to the gradient mismatch problem. In this study, we propose to address the gradient mismatch problem by designing a multi-branch ensemble model consisting of many weak hash code learners. Specifically, our design aggregates the gradients from multiple branches, which allows a better approximation of the gradients and regularizes the network. Our model adds little computational cost to the baseline BNN since a vast amount of network parameters are shared between the weak learners. Combining the efficiency of the BNNs and hash code learning, we obtain an effective ensemble model which is efficient both in feature extraction and ranking phases. Our experiments demonstrate that the proposed model outperforms a single BNN by more than %20 using nearly the same amount of floating point operations. Moreover, the proposed model outperforms a conventional ensemble of BNN by more than %7 while being nearly 10x and 2x more efficient in terms of CPU consumption and memory footprint, respectively. We explore the performance of BNNs for efficient person ReID as one of the first systems available in the literature. Moreover, we adopt the proposed ensemble model for further validation of the image classification task and observe that our method effectively regularizes BNNs, providing robustness to hyperparameter selection and producing more consistent results under different settings. |
first_indexed | 2024-03-13T04:18:04Z |
format | Article |
id | doaj.art-b9a97df0107e4c0493361543423b2181 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-03-13T04:18:04Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-b9a97df0107e4c0493361543423b21812023-06-20T23:00:26ZengIEEEIEEE Access2169-35362023-01-0111594465945510.1109/ACCESS.2023.328173710138817Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-IdentificationAyse Serbetci0https://orcid.org/0000-0003-4039-7244Yusuf Sinan Akgul1https://orcid.org/0000-0001-8501-4812Department of Computer Engineering, Istanbul Commerce University, Istanbul, TurkeyDepartment of Computer Engineering, Gebze Technical University, Kocaeli, TurkeyThis study aims to leverage Binary Neural Networks (BNN) to learn binary hash codes for efficient person re-identification (ReID). BNNs, which use binary weights and activations, show promise in speeding up the inference time in deep models. However, BNNs typically suffer from performance degradation mainly due to the discontinuity of the binarization operation. Proxy functions have been proposed to calculate the gradients in the backward propagation, but they lead to the gradient mismatch problem. In this study, we propose to address the gradient mismatch problem by designing a multi-branch ensemble model consisting of many weak hash code learners. Specifically, our design aggregates the gradients from multiple branches, which allows a better approximation of the gradients and regularizes the network. Our model adds little computational cost to the baseline BNN since a vast amount of network parameters are shared between the weak learners. Combining the efficiency of the BNNs and hash code learning, we obtain an effective ensemble model which is efficient both in feature extraction and ranking phases. Our experiments demonstrate that the proposed model outperforms a single BNN by more than %20 using nearly the same amount of floating point operations. Moreover, the proposed model outperforms a conventional ensemble of BNN by more than %7 while being nearly 10x and 2x more efficient in terms of CPU consumption and memory footprint, respectively. We explore the performance of BNNs for efficient person ReID as one of the first systems available in the literature. Moreover, we adopt the proposed ensemble model for further validation of the image classification task and observe that our method effectively regularizes BNNs, providing robustness to hyperparameter selection and producing more consistent results under different settings.https://ieeexplore.ieee.org/document/10138817/Binary neural networknetwork regularizationhash retrievalensemble learningperson re-identification |
spellingShingle | Ayse Serbetci Yusuf Sinan Akgul Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-Identification IEEE Access Binary neural network network regularization hash retrieval ensemble learning person re-identification |
title | Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-Identification |
title_full | Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-Identification |
title_fullStr | Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-Identification |
title_full_unstemmed | Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-Identification |
title_short | Regularizing Binary Neural Networks via Ensembling for Efficient Person Re-Identification |
title_sort | regularizing binary neural networks via ensembling for efficient person re identification |
topic | Binary neural network network regularization hash retrieval ensemble learning person re-identification |
url | https://ieeexplore.ieee.org/document/10138817/ |
work_keys_str_mv | AT ayseserbetci regularizingbinaryneuralnetworksviaensemblingforefficientpersonreidentification AT yusufsinanakgul regularizingbinaryneuralnetworksviaensemblingforefficientpersonreidentification |