{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,8,23]],"date-time":"2025-08-23T15:10:02Z","timestamp":1755961802006,"version":"3.44.0"},"reference-count":50,"publisher":"Association for Computing Machinery (ACM)","issue":"3","funder":[{"DOI":"10.13039\/501100004731","name":"Zhejiang Provincial Natural Science Foundation","doi-asserted-by":"crossref","award":["LDQ23F020001"],"award-info":[{"award-number":["LDQ23F020001"]}],"id":[{"id":"10.13039\/501100004731","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["62406286, 62072406"],"award-info":[{"award-number":["62406286, 62072406"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Key Laboratory of Data Protection and Intelligent Management, Ministry of Education, Sichuan University","award":["SCUSAKFKT202402Z"],"award-info":[{"award-number":["SCUSAKFKT202402Z"]}]},{"name":"Beijing Life Science Academy","award":["2024200CD0210"],"award-info":[{"award-number":["2024200CD0210"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Priv. Secur."],"published-print":{"date-parts":[[2025,8,31]]},"abstract":"<jats:p>\n            The increasing complexity of deep neural networks (DNNs) poses significant resource challenges for edge devices, prompting the development of compression technologies like model quantization. However, while improving model efficiency, quantization can introduce or perpetuate the original model\u2019s bias. Existing debiasing methods for quantized models often incur additional costs. To address this issue, we propose\n            <jats:italic toggle=\"yes\">FairQuanti<\/jats:italic>\n            , a novel quantization approach that leverages neuron role contribution to achieve fairness. By distinguishing between biased and normal neurons,\n            <jats:italic toggle=\"yes\">FairQuanti<\/jats:italic>\n            employs mixed precision quantization to mitigate model bias during the quantization process.\n            <jats:italic toggle=\"yes\">FairQuanti<\/jats:italic>\n            has four key differences from previous studies: (1)\u00a0\n            <jats:italic toggle=\"yes\">Neuron Roles<\/jats:italic>\n            - It formally defines biased and normal neuron roles, establishing a framework for feasible model quantization and bias mitigation; (2)\u00a0\n            <jats:italic toggle=\"yes\">Effectiveness<\/jats:italic>\n            - It introduces a fair quantization strategy that discriminatively quantizes neuron roles, balancing model accuracy and fairness through Bayesian optimization; (3)\u00a0\n            <jats:italic toggle=\"yes\">Generality<\/jats:italic>\n            - It applies to both structured and unstructured data across various quantization bit levels; (4)\u00a0\n            <jats:italic toggle=\"yes\">Robustness<\/jats:italic>\n            - It demonstrates resilience against adaptive attacks. Extensive experiments on five datasets (three structured and two unstructured) using five different models validate\n            <jats:italic toggle=\"yes\">FairQuanti<\/jats:italic>\n            \u2019s superior performance against eight baseline methods. Specifically, fairness metrics such as demographic parity (DP) improve by approximately 1.03 times, and the demographic parity ratio (DPR) improves by approximately 1.51 times compared to the baselines, with an average accuracy loss of less than 7.5% at 8-bit quantization.\n            <jats:italic toggle=\"yes\">FairQuanti<\/jats:italic>\n            presents a promising solution for deploying fair and efficient deep models on resource-constrained devices and holds potential for application in large language models to reduce size and computational demands while minimizing bias. Our source code is available at\n            <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" xlink:href=\"https:\/\/github.com\/Caozq2\/FairQuanti\">https:\/\/github.com\/Caozq2\/FairQuanti<\/jats:ext-link>\n            .\n          <\/jats:p>","DOI":"10.1145\/3744560","type":"journal-article","created":{"date-parts":[[2025,6,16]],"date-time":"2025-06-16T07:16:53Z","timestamp":1750058213000},"page":"1-28","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["FairQuanti: Enhancing Fairness in Deep Neural Network Quantization via Neuron Role Contribution"],"prefix":"10.1145","volume":"28","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7153-2755","authenticated-orcid":false,"given":"Jinyin","family":"Chen","sequence":"first","affiliation":[{"name":"Zhejiang University of Technology","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-4292-1817","authenticated-orcid":false,"given":"Zhiqi","family":"Cao","sequence":"additional","affiliation":[{"name":"Zhejiang University of Technology","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-5386-6395","authenticated-orcid":false,"given":"Xiaojuan","family":"Wang","sequence":"additional","affiliation":[{"name":"Zhejiang University of Technology","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8997-5343","authenticated-orcid":false,"given":"Haibin","family":"Zheng","sequence":"additional","affiliation":[{"name":"Zhejiang University of Technology","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6766-4579","authenticated-orcid":false,"given":"Zhaoyan","family":"Ming","sequence":"additional","affiliation":[{"name":"Hangzhou City University","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-2010-4898","authenticated-orcid":false,"given":"Yayu","family":"Zheng","sequence":"additional","affiliation":[{"name":"Zhejiang University of Technology","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2025,8,23]]},"reference":[{"key":"e_1_3_2_2_2","unstructured":"Oscar Andersson and William Isaksson. 2022. Investigating and mitigating effects of quantization on algorithmic bias. LU-CS-EX. 2022."},{"key":"e_1_3_2_3_2","doi-asserted-by":"publisher","DOI":"10.1147\/JRD.2019.2942287"},{"key":"e_1_3_2_4_2","unstructured":"Alex Beutel Jilin Chen Zhe Zhao and Ed H Chi. 2017. Data decisions and theoretical implications when adversarially learning fair representations. arXiv:1707.00075. Retrieved from https:\/\/arxiv.org\/abs\/1707.00075"},{"key":"e_1_3_2_5_2","first-page":"855","volume-title":"Proceedings of the 2022 26th International Conference on Pattern Recognition","author":"Boutros Fadi","year":"2022","unstructured":"Fadi Boutros, Naser Damer, and Arjan Kuijper. 2022. Quantface: Towards lightweight face recognition by synthetic data low-bit quantization. In Proceedings of the 2022 26th International Conference on Pattern Recognition. IEEE, 855\u2013862."},{"key":"e_1_3_2_6_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01318"},{"key":"e_1_3_2_7_2","unstructured":"Junyi Chai and Xiaoqian Wang. 2023. To be robust and to be fair: Aligning fairness with robustness. arXiv:2304.00061. Retrieved from https:\/\/arxiv.org\/abs\/2304.00061"},{"key":"e_1_3_2_8_2","article-title":"Rsmamba: Remote sensing image classification with state space model","author":"Chen Keyan","year":"2024","unstructured":"Keyan Chen, Bowen Chen, Chenyang Liu, Wenyuan Li, Zhengxia Zou, and Zhenwei Shi. 2024. Rsmamba: Remote sensing image classification with state space model. IEEE Geoscience and Remote Sensing Letters 21, 1 (2024), 1\u20135.","journal-title":"IEEE Geoscience and Remote Sensing Letters"},{"key":"e_1_3_2_9_2","doi-asserted-by":"publisher","DOI":"10.1145\/3663363"},{"key":"e_1_3_2_10_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2024.3447085"},{"key":"e_1_3_2_11_2","first-page":"1887","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Choi Kristy","year":"2020","unstructured":"Kristy Choi, Aditya Grover, Trisha Singh, Rui Shu, and Stefano Ermon. 2020. Fair generative modeling via weak supervision. In Proceedings of the International Conference on Machine Learning. PMLR, 1887\u20131898."},{"key":"e_1_3_2_12_2","article-title":"Deep learning model compression with rank reduction in tensor decomposition","author":"Dai Wei","year":"2023","unstructured":"Wei Dai, Jicong Fan, Yiming Miao, and Kai Hwang. 2023. Deep learning model compression with rank reduction in tensor decomposition. IEEE Transactions on Neural Networks and Learning Systems 36, 1 (2023), 1315\u20131328.","journal-title":"IEEE Transactions on Neural Networks and Learning Systems"},{"key":"e_1_3_2_13_2","doi-asserted-by":"publisher","DOI":"10.1145\/3637528.3671474"},{"key":"e_1_3_2_14_2","doi-asserted-by":"publisher","DOI":"10.1145\/3468507.3468513"},{"key":"e_1_3_2_15_2","first-page":"5513","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Ding Xiaohan","year":"2024","unstructured":"Xiaohan Ding, Yiyuan Zhang, Yixiao Ge, Sijie Zhao, Lin Song, Xiangyu Yue, and Ying Shan. 2024. UniRepLKNet: A universal perception large-kernel convnet for audio video point cloud time-series and image recognition. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. 5513\u20135524."},{"key":"e_1_3_2_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/2090236.2090255"},{"key":"e_1_3_2_17_2","doi-asserted-by":"crossref","first-page":"69","DOI":"10.1007\/978-3-030-58536-5_5","volume-title":"Computer Vision\u2013ECCV 2020: 16th European Conference, Glasgow, UK, August 23\u201328, 2020, Proceedings, Part II 16","author":"Fang Jun","year":"2020","unstructured":"Jun Fang, Ali Shafiee, Hamzah Abdel-Aziz, David Thorsley, Georgios Georgiadis, and Joseph H. Hassoun. 2020. Post-training piecewise linear quantization for deep neural networks. In Computer Vision\u2013ECCV 2020: 16th European Conference, Glasgow, UK, August 23\u201328, 2020, Proceedings, Part II 16. Springer, 69\u201386."},{"key":"e_1_3_2_18_2","doi-asserted-by":"publisher","DOI":"10.1145\/3510003.3510087"},{"key":"e_1_3_2_19_2","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-021-01453-z"},{"key":"e_1_3_2_20_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"issue":"241","key":"e_1_3_2_21_2","first-page":"1","article-title":"Sparsity in deep learning: Pruning and growth for efficient inference and training in neural networks","volume":"22","author":"Hoefler Torsten","year":"2021","unstructured":"Torsten Hoefler, Dan Alistarh, Tal Ben-Nun, Nikoli Dryden, and Alexandra Peste. 2021. Sparsity in deep learning: Pruning and growth for efficient inference and training in neural networks. Journal of Machine Learning Research 22, 241 (2021), 1\u2013124.","journal-title":"Journal of Machine Learning Research"},{"key":"e_1_3_2_22_2","unstructured":"Sara Hooker Nyalleng Moorosi Gregory Clark Samy Bengio and Emily Denton. 2020. Characterising bias in compressed models. arXiv:2010.03058. Retrieved from https:\/\/arxiv.org\/abs\/2010.03058"},{"key":"e_1_3_2_23_2","unstructured":"Xijie Huang Zhiqiang Shen and Kwang-Ting Cheng. 2023. Variation-aware vision transformer quantization. arXiv:2307.00331. Retrieved from https:\/\/arxiv.org\/abs\/2307.00331"},{"key":"e_1_3_2_24_2","first-page":"12115","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Jung Sangwon","year":"2021","unstructured":"Sangwon Jung, Donggyu Lee, Taeeon Park, and Taesup Moon. 2021. Fair feature distillation for visual recognition. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. 12115\u201312124."},{"key":"e_1_3_2_25_2","volume-title":"Proceedings of naacL-HLT","author":"Kenton Jacob Devlin Ming-Wei Chang","year":"2019","unstructured":"Jacob Devlin Ming-Wei Chang Kenton and Lee Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of naacL-HLT. Minneapolis, Minnesota."},{"key":"e_1_3_2_26_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.engappai.2024.109114"},{"key":"e_1_3_2_27_2","first-page":"1","volume-title":"Proceedings of the 46th IEEE\/ACM International Conference on Software Engineering","author":"Li Tianlin","year":"2024","unstructured":"Tianlin Li, Yue Cao, Jian Zhang, Shiqian Zhao, Yihao Huang, Aishan Liu, Qing Guo, and Yang Liu. 2024. RUNNER: Responsible UNfair NEuron repair for enhancing deep neural network fairness. In Proceedings of the 46th IEEE\/ACM International Conference on Software Engineering. 1\u201313."},{"key":"e_1_3_2_28_2","unstructured":"Yuhang Li Ruihao Gong Xu Tan Yang Yang Peng Hu Qi Zhang Fengwei Yu Wei Wang and Shi Gu. 2021. Brecq: Pushing the limit of post-training quantization by block reconstruction. arXiv:2102.05426. Retrieved from https:\/\/arxiv.org\/abs\/2102.05426"},{"key":"e_1_3_2_29_2","unstructured":"Yinhan Liu. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv:1907.11692. Retrieved from https:\/\/arxiv.org\/abs\/1907.11692"},{"key":"e_1_3_2_30_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.425"},{"key":"e_1_3_2_31_2","unstructured":"Yuzhen Mao Zhun Deng Huaxiu Yao Ting Ye Kenji Kawaguchi and James Zou. 2023. Last-layer fairness fine-tuning is simple and effective for neural networks. arXiv:2304.03935. Retrieved from https:\/\/arxiv.org\/abs\/2304.03935"},{"key":"e_1_3_2_32_2","doi-asserted-by":"publisher","DOI":"10.1145\/3457607"},{"key":"e_1_3_2_33_2","first-page":"15720","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Miles Roy","year":"2024","unstructured":"Roy Miles, Ismail Elezi, and Jiankang Deng. 2024. VkD: Improving knowledge distillation using orthogonal projections. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. 15720\u201315730."},{"key":"e_1_3_2_34_2","first-page":"1611","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Misra Diganta","year":"2024","unstructured":"Diganta Misra, Muawiz Chaudhary, Agam Goyal, Bharat Runwal, and Pin Yu Chen. 2024. Uncovering the hidden cost of model compression. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. 1611\u20131621."},{"key":"e_1_3_2_35_2","first-page":"7197","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Nagel Markus","year":"2020","unstructured":"Markus Nagel, Rana Ali Amjad, Mart Van Baalen, Christos Louizos, and Tijmen Blankevoort. 2020. Up or down? Adaptive rounding for post-training quantization. In Proceedings of the International Conference on Machine Learning. PMLR, 7197\u20137206."},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00141"},{"key":"e_1_3_2_37_2","first-page":"1","volume-title":"Proceedings of the 2023 International Conference of the Biometrics Special Interest Group","author":"Neto Pedro C.","year":"2023","unstructured":"Pedro C. Neto, Eduarda Caldeira, Jaime S. Cardoso, and Ana F. Sequeira. 2023. Compressed models decompress race biases: What quantized models forget for fair face recognition. In Proceedings of the 2023 International Conference of the Biometrics Special Interest Group. IEEE, 1\u20135."},{"key":"e_1_3_2_38_2","first-page":"6668","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Parihar Rishubh","year":"2024","unstructured":"Rishubh Parihar, Abhijnya Bhat, Abhipsa Basu, Saswat Mallick, Jogendra Nath Kundu, and R. Venkatesh Babu. 2024. Balancing act: Distribution-guided debiasing in diffusion models. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. 6668\u20136678."},{"key":"e_1_3_2_39_2","unstructured":"Rakesh Pavan Andrew Lowy Sina Baharlouei Meisam Razaviyayn and Ahmad Beirami. 2020. FERMI: Fair empirical risk minimization via exponential r\u00e9nyi mutual information. arXiv:2304.03935. Retrieved from https:\/\/arxiv.org\/abs\/2304.03935"},{"key":"e_1_3_2_40_2","first-page":"20311","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Qraitem Maan","year":"2023","unstructured":"Maan Qraitem, Kate Saenko, and Bryan A. Plummer. 2023. Bias mimicking: A simple sampling approach for bias mitigation. In Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition. 20311\u201320320."},{"key":"e_1_3_2_41_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00918"},{"key":"e_1_3_2_42_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00474"},{"key":"e_1_3_2_43_2","doi-asserted-by":"crossref","DOI":"10.1109\/TPAMI.2024.3450575","article-title":"DeepTensor: Low-rank tensor decomposition with deep network priors","author":"Saragadam Vishwanath","year":"2024","unstructured":"Vishwanath Saragadam, Randall Balestriero, Ashok Veeraraghavan, and Richard G Baraniuk. 2024. DeepTensor: Low-rank tensor decomposition with deep network priors. IEEE Transactions on Pattern Analysis and Machine Intelligence 46, 12 (2024), 10337\u201310348.","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"e_1_3_2_44_2","first-page":"121","volume-title":"Proceedings of the International Conference on Pattern Recognition","author":"Stoychev Samuil","year":"2022","unstructured":"Samuil Stoychev and Hatice Gunes. 2022. The effect of model compression on fairness in facial expression recognition. In Proceedings of the International Conference on Pattern Recognition. Springer, 121\u2013138."},{"key":"e_1_3_2_45_2","doi-asserted-by":"crossref","unstructured":"Siqi Sun Yu Cheng Zhe Gan and Jingjing Liu. 2019. Patient knowledge distillation for bert model compression. arXiv:1908.09355. Retrieved from https:\/\/arxiv.org\/abs\/1908.09355","DOI":"10.18653\/v1\/D19-1441"},{"key":"e_1_3_2_46_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.463"},{"key":"e_1_3_2_47_2","doi-asserted-by":"crossref","unstructured":"Jieyu Zhao Tianlu Wang Mark Yatskar Vicente Ordonez and Kai-Wei Chang. 2018. Gender bias in coreference resolution: Evaluation and debiasing methods. arXiv:1804.06876. Retrieved from https:\/\/arxiv.org\/abs\/1804.06876","DOI":"10.18653\/v1\/N18-2003"},{"key":"e_1_3_2_48_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2024.3392335"},{"key":"e_1_3_2_49_2","doi-asserted-by":"crossref","first-page":"1519","DOI":"10.1145\/3510003.3510123","volume-title":"Proceedings of the 44th International Conference on Software Engineering","author":"Zheng Haibin","year":"2022","unstructured":"Haibin Zheng, Zhiqing Chen, Tianyu Du, Xuhong Zhang, Yao Cheng, Shouling Ji, Jingyi Wang, Yue Yu, and Jinyin Chen. 2022. Neuronfair: Interpretable white-box fairness testing through biased neuron identification. In Proceedings of the 44th International Conference on Software Engineering. 1519\u20131531."},{"issue":"1","key":"e_1_3_2_50_2","doi-asserted-by":"crossref","first-page":"62","DOI":"10.54097\/v160aa61","article-title":"Medication recommendation system based on natural language processing for patient emotion analysis","volume":"10","author":"Zheng Haotian","year":"2024","unstructured":"Haotian Zheng, Kangming Xu, Huiming Zhou, Yufu Wang, and Guangze Su. 2024. Medication recommendation system based on natural language processing for patient emotion analysis. Academic Journal of Science and Technology 10, 1 (2024), 62\u201368.","journal-title":"Academic Journal of Science and Technology"},{"key":"e_1_3_2_51_2","unstructured":"Michael Zhu and Suyog Gupta. 2017. To prune or not to prune: Exploring the efficacy of pruning for model compression. arXiv:1710.01878. Retrieved from https:\/\/arxiv.org\/abs\/1710.01878"}],"container-title":["ACM Transactions on Privacy and Security"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3744560","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,23]],"date-time":"2025-08-23T14:48:10Z","timestamp":1755960490000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3744560"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,23]]},"references-count":50,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2025,8,31]]}},"alternative-id":["10.1145\/3744560"],"URL":"https:\/\/doi.org\/10.1145\/3744560","relation":{},"ISSN":["2471-2566","2471-2574"],"issn-type":[{"type":"print","value":"2471-2566"},{"type":"electronic","value":"2471-2574"}],"subject":[],"published":{"date-parts":[[2025,8,23]]},"assertion":[{"value":"2024-10-22","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-05-04","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-08-23","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}