QuantNAS for super resolution: searching for efficient quantization-friendly architectures against quantization noise

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shvetsov, Egor, Osin, Dmitry, Zaytsev, Alexey, Koryakovskiy, Ivan, Buchnev, Valentin, Trofimov, Ilya, Burnaev, Evgeny
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916086329376768
author Shvetsov, Egor
Osin, Dmitry
Zaytsev, Alexey
Koryakovskiy, Ivan
Buchnev, Valentin
Trofimov, Ilya
Burnaev, Evgeny
author_facet Shvetsov, Egor
Osin, Dmitry
Zaytsev, Alexey
Koryakovskiy, Ivan
Buchnev, Valentin
Trofimov, Ilya
Burnaev, Evgeny
contents There is a constant need for high-performing and computationally efficient neural network models for image super-resolution: computationally efficient models can be used via low-capacity devices and reduce carbon footprints. One way to obtain such models is to compress models, e.g. quantization. Another way is a neural architecture search that automatically discovers new, more efficient solutions. We propose a novel quantization-aware procedure, the QuantNAS that combines pros of these two approaches. To make QuantNAS work, the procedure looks for quantization-friendly super-resolution models. The approach utilizes entropy regularization, quantization noise, and Adaptive Deviation for Quantization (ADQ) module to enhance the search procedure. The entropy regularization technique prioritizes a single operation within each block of the search space. Adding quantization noise to parameters and activations approximates model degradation after quantization, resulting in a more quantization-friendly architectures. ADQ helps to alleviate problems caused by Batch Norm blocks in super-resolution models. Our experimental results show that the proposed approximations are better for search procedure than direct model quantization. QuantNAS discovers architectures with better PSNR/BitOps trade-off than uniform or mixed precision quantization of fixed architectures. We showcase the effectiveness of our method through its application to two search spaces inspired by the state-of-the-art SR models and RFDN. Thus, anyone can design a proper search space based on an existing architecture and apply our method to obtain better quality and efficiency. The proposed procedure is 30\% faster than direct weight quantization and is more stable.
format Preprint
id arxiv_https___arxiv_org_abs_2208_14839
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle QuantNAS for super resolution: searching for efficient quantization-friendly architectures against quantization noise
Shvetsov, Egor
Osin, Dmitry
Zaytsev, Alexey
Koryakovskiy, Ivan
Buchnev, Valentin
Trofimov, Ilya
Burnaev, Evgeny
Computer Vision and Pattern Recognition
There is a constant need for high-performing and computationally efficient neural network models for image super-resolution: computationally efficient models can be used via low-capacity devices and reduce carbon footprints. One way to obtain such models is to compress models, e.g. quantization. Another way is a neural architecture search that automatically discovers new, more efficient solutions. We propose a novel quantization-aware procedure, the QuantNAS that combines pros of these two approaches. To make QuantNAS work, the procedure looks for quantization-friendly super-resolution models. The approach utilizes entropy regularization, quantization noise, and Adaptive Deviation for Quantization (ADQ) module to enhance the search procedure. The entropy regularization technique prioritizes a single operation within each block of the search space. Adding quantization noise to parameters and activations approximates model degradation after quantization, resulting in a more quantization-friendly architectures. ADQ helps to alleviate problems caused by Batch Norm blocks in super-resolution models. Our experimental results show that the proposed approximations are better for search procedure than direct model quantization. QuantNAS discovers architectures with better PSNR/BitOps trade-off than uniform or mixed precision quantization of fixed architectures. We showcase the effectiveness of our method through its application to two search spaces inspired by the state-of-the-art SR models and RFDN. Thus, anyone can design a proper search space based on an existing architecture and apply our method to obtain better quality and efficiency. The proposed procedure is 30\% faster than direct weight quantization and is more stable.
title QuantNAS for super resolution: searching for efficient quantization-friendly architectures against quantization noise
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2208.14839