paper-with-me

홈 › Papers

FQ-Conv: Fully Quantized Convolution for Efficient and Accurate Inference

2019-12-19 · Bram-Ernst Verhoef, Nathan Laubeuf, Stefan Cosemans, Peter Debacker, Ioannis Papistas, Arindam Mallik, Diederik Verkest

Deep neural networks (DNNs) can be made hardware-efficient by reducing the numerical precision of the weights and activations of the network and by improving the network's resilience to noise. However, this gain in efficiency often comes at the cost of significantly reduced accuracy. In this paper, we present a novel approach to quantizing convolutional neural network. The resulting networks perform all computations in low-precision, without requiring higher-precision BN and nonlinearities, while still being highly accurate. To achieve this result, we employ a novel quantization technique that learns to optimally quantize the weights and activations of the network during training. Additionally, to enhance training convergence we use a new training technique, called gradual quantization. We leverage the nonlinear and normalizing behavior of our quantization function to effectively remove the higher-precision nonlinearities and BN from the network. The resulting convolutional layers are fully quantized to low precision, from input to output, ideal for neural network accelerators on the edge. We demonstrate the potential of this approach on different datasets and networks, showing that ternary-weight CNNs with low-precision in- and outputs perform virtually on par with their full-precision equivalents. Finally, we analyze the influence of noise on the weights, activations and convolution outputs (multiply-accumulate, MAC) and propose a strategy to improve network performance under noisy conditions.

📄 PDF Abstract BibTeX arXiv:1912.09356

Code (0)

등록된 구현이 없습니다.

Tasks

Quantization

Methods 이 논문이 사용한 방법론

Convolution A convolution is a type of matrix operation, consisting of a kernel, a small matrix of weights, that slides over input data performing element-wise multiplication with the…

Similar Papers 제목 키워드 기반

AQD: Towards Accurate Fully-Quantized Object Detection

2020-07-14 · CVPR 2021 1 · Peng Chen, Jing Liu, Bohan Zhuang, Mingkui Tan 외

Network quantization allows inference to be conducted using low-precision arithmetic for improved inference efficiency of deep neural networks on edge devices. However, designing aggressively low-bit (e.g., 2-bit) quanti…

Image ClassificationObjectobject-detectionObject Detection+1

Quantized Convolutional Neural Networks for Mobile Devices

2015-12-21 · CVPR 2016 6 · Jiaxiang Wu, Cong Leng, Yuhang Wang, Qinghao Hu 외

Recently, convolutional neural networks (CNN) have demonstrated impressive performance in various computer vision tasks. However, high performance hardware is typically indispensable for the application of CNN models due…

General Classification

SeerNet: Predicting Convolutional Neural Network Feature-Map Sparsity Through Low-Bit Quantization

2019-06-01 · CVPR 2019 6 · Shijie Cao, Lingxiao Ma, Wencong Xiao, Chen Zhang 외

In this paper we present a novel and general method to accelerate convolutional neural network (CNN) inference by taking advantage of feature map sparsity. We experimentally demonstrate that a highly quantized version of…

Quantization

FQ-ViT: Post-Training Quantization for Fully Quantized Vision Transformer

2021-11-27 · Yang Lin, Tianyu Zhang, Peiqin Sun, Zheng Li 외

Network quantization significantly reduces model inference complexity and has been widely used in real-world deployments. However, most existing quantization methods have been developed mainly on Convolutional Neural Net…

Quantization

Data-Free Group-Wise Fully Quantized Winograd Convolution via Learnable Scales

2024-12-27 · CVPR 2025 1 · Shuokai Pan, Gerti Tuzi, Sudarshan Sreeram, Dibakar Gope

Despite the revolutionary breakthroughs of large-scale text-to-image diffusion models for complex vision and downstream tasks, their extremely high computational and storage costs limit their usability. Quantization of d…

image-classificationImage ClassificationImage GenerationQuantization+2