{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/hptq-hardware-friendly-post-training","title":"HPTQ: Hardware-Friendly Post Training Quantization","arxiv_id":"2109.09113","date":"2021-09-19","proceeding":null,"authors":["Hai Victor Habi","Reuven Peretz","Elad Cohen","Lior Dikstein","Oranit Dror","Idit Diamant","Roy H. Jennings","Arnon Netzer"],"abstract":"Neural network quantization enables the deployment of models on edge devices. An essential requirement for their hardware efficiency is that the quantizers are hardware-friendly: uniform, symmetric, and with power-of-two thresholds. To the best of our knowledge, current post-training quantization methods do not support all of these constraints simultaneously. In this work, we introduce a hardware-friendly post training quantization (HPTQ) framework, which addresses this problem by synergistically combining several known quantization methods. We perform a large-scale study on four tasks: classification, object detection, semantic segmentation and pose estimation over a wide variety of network architectures. Our extensive experiments show that competitive results can be obtained under hardware-friendly constraints.","url_abs":"https://arxiv.org/abs/2109.09113v3","url_pdf":"https://arxiv.org/pdf/2109.09113v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"hptq-hardware-friendly-post-training","repo_url":"https://github.com/sony/model_optimization","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"},{"task_slug":"quantization","task_name":"Quantization"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/quantization-on-coco","task":"Quantization","dataset":"COCO (Common Objects in Context)","model":"SSD ResNet50 V1 FPN 640x640","rank_in_archive_order":1,"of":1,"metrics":{"MAP":"34.3"},"uses_additional_data":false},{"leaderboard":"/sota/quantization-on-imagenet","task":"Quantization","dataset":"ImageNet","model":"Xception W8A8","rank_in_archive_order":8,"of":27,"metrics":{"Activation bits":"8","Top-1 Accuracy (%)":"78.972","Weight bits":"8"},"uses_additional_data":false},{"leaderboard":"/sota/quantization-on-imagenet","task":"Quantization","dataset":"ImageNet","model":"EfficientNet-B0 ReLU W8A8","rank_in_archive_order":11,"of":27,"metrics":{"Activation bits":"8","Top-1 Accuracy (%)":"77.092","Weight bits":"8"},"uses_additional_data":false},{"leaderboard":"/sota/quantization-on-imagenet","task":"Quantization","dataset":"ImageNet","model":"EfficientNet-B0 W8A8","rank_in_archive_order":16,"of":27,"metrics":{"Activation bits":"8","Top-1 Accuracy (%)":"74.216","Weight bits":"8"},"uses_additional_data":false},{"leaderboard":"/sota/quantization-on-imagenet","task":"Quantization","dataset":"ImageNet","model":"DenseNet-121 W8A8","rank_in_archive_order":19,"of":27,"metrics":{"Activation bits":"8","Top-1 Accuracy (%)":"73.356","Weight bits":"8"},"uses_additional_data":false},{"leaderboard":"/sota/quantization-on-imagenet","task":"Quantization","dataset":"ImageNet","model":"MobileNetV2 W8A8","rank_in_archive_order":23,"of":27,"metrics":{"Activation bits":"8","Top-1 Accuracy (%)":"71.46","Weight bits":"8"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2109.09113","atlas_url":"https://app.syntology.ai/?focus=2109.09113","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}