MathWorks - Mobile View
  • 碻形冰暨硞 MathWorks 帐憷碻形冰暨硞 MathWorks 帐憷
  • Access your MathWorks Account
    • 我的帐户
    • 我的社区资料
    • 关联许可证
    • 登出
  • 产品
  • 解决方案
  • 学术
  • 支持
  • 社区
  • 活动
  • 获取 MATLAB
MathWorks
  • 产品
  • 解决方案
  • 学术
  • 支持
  • 社区
  • 活动
  • 获取 MATLAB
  • 碻形冰暨硞 MathWorks 帐憷碻形冰暨硞 MathWorks 帐憷
  • Access your MathWorks Account
    • 我的帐户
    • 我的社区资料
    • 关联许可证
    • 登出

视频与网上研讨会

  • MathWorks
  • 视频
  • 视频首页
  • 搜索
  • 视频首页
  • 搜索
  • 联系销售
  • 试用软件
  Register to watch video
  • Description
  • Full Transcript
  • Related Resources

Quantizing a Deep Learning Network in MATLAB

Ram Cherukuri, MathWorks

In this video, we demonstrate the deep learning quantization workflow in MATLAB. Using the Model Quantization Library Support Package, we illustrate how you can calibrate, quantize, and validate a deep learning network such as Resnet50. We also highlight the impact of quantization on reducing the memory of some standard networks such as Resnet101 and InceptionV3.

 

Deep Learning quantization is a key optimization strategy for efficient deployment of deep learning networks, particularly on embedded platforms.

I am Ram Cherukuri, senior product manager at MathWorks and in this video I will give you an overview of the deep learning quantization workflow in MATLAB.

Quantizing the weights, biases, and activations to lower precision data types like INT8 or FP16 significantly reduces the memory footprint of the AI algorithm and can result in improved inference performance on the embedded hardware.

You can use the Model Quantization Library Support Package for quantizing your deep learning network in MATLAB. You can download it from the Add-On Explorer as shown here.

The quantization workflow leverages instrumentation, based on a calibration datastore to compute the instrumentation statistics that are used to quantize the weights, biases, and activations of the layers of the network.

Finally, the validation step computes accuracy metrics to analyze and understand the impact of quantization on the accuracy of the network. Let’s take Resnet50 as an example network to go through this workflow.

Here is the Deep Learning Quantizer app, where you first import the network from the MATLAB workspace and you will see the network structure displayed on the left side pane.

Next, you select the data store that you would like to use for calibration and the app displays the computed statistics such as the min and max values of weights, biases, and activations of each layer. You can also choose the layers that you can quantize and then validate the impact of quantization using a validation datastore.

In this example, we have used the default top 1 accuracy metric and you can see that there is a 67% reduction in memory with no drop in accuracy.  You can then proceed to generate code from the quantized network for deployment.

We repeated this workflow with a few networks, only quantizing the compute-intensive conv layers to INT8.

You can see the impact of quantization in the chart here. For instance, the largest network here with 180 MB in memory, Resnet101, sees 72% compression with 2% drop in accuracy. InceptionV3, on the other hand, has the largest drop in accuracy of 4%, with 67% compression, going from 100 MB to 33 MB in memory.

This highlights the significant impact of quantization for efficient deployment of deep learning networks.

Please refer to the resources below the video to learn how to get started and explore these new capabilities in MATLAB.

Related Products

  • Deep Learning Toolbox
  • GPU Coder

Learn More

Download the support package
What Is int8 Quantization and Why Is It Popular for Deep Neural Networks?
INT8 Quantization with Deep Network Quantizer
Quantization of Deep Neural Networks
Related Information
Try out the Deep Network Quantizer app

Feedback

Featured Product

Deep Learning Toolbox

  • Request Trial
  • Get Pricing

Up Next:

24:56
Optimal Neural Network for Automotive Product Development

Related Videos:

7:35
Deep Learning for Computer Vision with MATLAB (Highlights)
29:45
CAN Communication with Your ECUs and the Vehicle Network...
27:59
Deep Learning for Computer Vision
46:41
Aircraft Power Network Development with Model-Based Design

View more related videos

MathWorks - Domain Selector

Select a Web Site

Choose a web site to get translated content where available and see local events and offers. Based on your location, we recommend that you select: .

Select web site

You can also select a web site from the following list:

How to Get Best Site Performance

Select the China site (in Chinese or English) for best site performance. Other MathWorks country sites are not optimized for visits from your location.

Americas

  • América Latina (Español)
  • Canada (English)
  • United States (English)

Europe

  • Belgium (English)
  • Denmark (English)
  • Deutschland (Deutsch)
  • España (Español)
  • Finland (English)
  • France (Français)
  • Ireland (English)
  • Italia (Italiano)
  • Luxembourg (English)
  • Netherlands (English)
  • Norway (English)
  • Österreich (Deutsch)
  • Portugal (English)
  • Sweden (English)
  • Switzerland
    • Deutsch
    • English
    • Français
  • United Kingdom (English)

Asia Pacific

  • Australia (English)
  • India (English)
  • New Zealand (English)
  • 中国
    • 简体中文Chinese
    • English
  • 日本Japanese (日本語)
  • 한국Korean (한국어)

Contact your local office

  • 联系销售
  • 试用软件

了解产品

  • MATLAB
  • Simulink
  • 学生版软件
  • 硬件支持
  • 文件交换

试用或购买

  • 下载
  • 试用软件
  • 联系销售
  • 定价和许可
  • 如何购买

如何使用

  • 文档
  • 教程
  • 示例
  • 视频与网上研讨会
  • 培训

获取支持

  • 安装帮助
  • MATLAB 问答社区
  • 咨询
  • 许可中心
  • 联系支持

关于 MathWorks

  • 招聘
  • 新闻室
  • 社会愿景
  • 联系销售
  • 关于 MathWorks

MathWorks

Accelerating the pace of engineering and science

MathWorks 公司是世界领先的为工程师和科学家提供数学计算软件的开发商。

发现…

  • Select a Web Site United States
  • 专利
  • 商标
  • 隐私权政策
  • 防盗版
  • 应用状态

京ICP备12052471号

© 1994-2021 The MathWorks, Inc.

  • Weibo
  • WeChat

    WeChat

  • Bilibili
  • Youku
  • Facebook
  • Twitter
  • LinkedIn
  • RSS

关注我们