TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Papers/HPTQ: Hardware-Friendly Post Training Quantization

HPTQ: Hardware-Friendly Post Training Quantization

Hai Victor Habi, Reuven Peretz, Elad Cohen, Lior Dikstein, Oranit Dror, Idit Diamant, Roy H. Jennings, Arnon Netzer

2021-09-19QuantizationSemantic SegmentationPose Estimationobject-detectionObject Detection
PaperPDFCode(official)

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.

Results

TaskDatasetMetricValueModel
QuantizationImageNetActivation bits8Xception W8A8
QuantizationImageNetTop-1 Accuracy (%)78.972Xception W8A8
QuantizationImageNetWeight bits8Xception W8A8
QuantizationImageNetActivation bits8EfficientNet-B0 ReLU W8A8
QuantizationImageNetTop-1 Accuracy (%)77.092EfficientNet-B0 ReLU W8A8
QuantizationImageNetWeight bits8EfficientNet-B0 ReLU W8A8
QuantizationImageNetActivation bits8EfficientNet-B0 W8A8
QuantizationImageNetTop-1 Accuracy (%)74.216EfficientNet-B0 W8A8
QuantizationImageNetWeight bits8EfficientNet-B0 W8A8
QuantizationImageNetActivation bits8DenseNet-121 W8A8
QuantizationImageNetTop-1 Accuracy (%)73.356DenseNet-121 W8A8
QuantizationImageNetWeight bits8DenseNet-121 W8A8
QuantizationImageNetActivation bits8MobileNetV2 W8A8
QuantizationImageNetTop-1 Accuracy (%)71.46MobileNetV2 W8A8
QuantizationImageNetWeight bits8MobileNetV2 W8A8
QuantizationCOCO (Common Objects in Context)MAP34.3SSD ResNet50 V1 FPN 640x640

Related Papers

Efficient Deployment of Spiking Neural Networks on SpiNNaker2 for DVS Gesture Recognition Using Neuromorphic Intermediate Representation2025-09-04SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction2025-07-21An End-to-End DNN Inference Framework for the SpiNNaker2 Neuromorphic MPSoC2025-07-18Task-Specific Audio Coding for Machines: Machine-Learned Latent Features Are Codes for That Machine2025-07-17Angle Estimation of a Single Source with Massive Uniform Circular Arrays2025-07-17DiffOSeg: Omni Medical Image Segmentation via Multi-Expert Collaboration Diffusion Model2025-07-17SCORE: Scene Context Matters in Open-Vocabulary Remote Sensing Instance Segmentation2025-07-17Unified Medical Image Segmentation with State Space Modeling Snake2025-07-17