English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep neural networks (DNNs) have demonstrated their effectiveness in a wide range of computer vision tasks, with the state-of-the-art results obtained through complex and deep structures that require intensive computation and memory. In the past, graphic processing units enabled these breakthroughs because of their greater computational speed. Now-a-days, efficient model inference is crucial for consumer applications on resource-constrained platforms. As a result, there is much interest in the research and development of dedicated deep learning (DL) hardware to improve the throughput and energy efficiency of DNNs. Low-precision representation of DNN data-structures through quantization would bring great benefits to specialized DL hardware especially when expensive floating-point operations can be avoided and replaced by more efficient fixed-point operations. However, the rigorous quantization leads to a severe accuracy drop. As such, quantization opens a large hyper-parameter space at bit-precision levels, the exploration of which is a major challenge. In this paper, we propose a novel framework referred to as the <bold>F</bold>i<bold>x</bold>ed-<bold>P</bold>oint <bold>Q</bold>uantizer of deep neural <bold>Net</bold>works (<bold>FxP-QNet</bold>) that flexibly designs a mixed low-precision DNN for integer-arithmetic-only deployment. Specifically, the <bold>FxP-QNet</bold> gradually adapts the quantization level for each data-structure of each layer based on the trade-off between the network accuracy and the low-precision requirements. Additionally, it employs post-training self-distillation and network prediction error statistics to optimize the quantization of floating-point values into fixed-point numbers. Examining <bold>FxP-QNet</bold> on state-of-the-art architectures and the benchmark ImageNet dataset, we empirically demonstrate the effectiveness of <bold>FxP-QNet</bold> in achieving the accuracy-compression trade-off without the need for training. The results show that <bold>FxP-QNet</bold>-quantized AlexNet, VGG-16, and ResNet-18 reduce the overall memory requirements of their full-precision counterparts by <inline-formula> <tex-math notation="LaTeX">$7.16\times $ </tex-math></inline-formula>, <inline-formula> <tex-math notation="LaTeX">$10.36\times $ </tex-math></inline-formula>, and <inline-formula> <tex-math notation="LaTeX">$6.44\times $ </tex-math></inline-formula> with less than 0.95&#x0025;, 0.95&#x0025;, and 1.99&#x0025; accuracy drop, respectively.

FxP-QNet: A Post-Training Quantizer for the Design of Mixed Low-Precision DNNs With Dynamic Fixed-Point Representation