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.
@article{arxiv.2109.09113,
title = {HPTQ: Hardware-Friendly Post Training Quantization},
author = {Hai Victor Habi and Reuven Peretz and Elad Cohen and Lior Dikstein and Oranit Dror and Idit Diamant and Roy H. Jennings and Arnon Netzer},
journal= {arXiv preprint arXiv:2109.09113},
year = {2021}
}