Deep Learning For Vision Analytics. SAS User Group Malaysia 3 rd May, 2018

Size: px
Start display at page:

Download "Deep Learning For Vision Analytics. SAS User Group Malaysia 3 rd May, 2018"

Transcription

1 Deep Learning For Vision Analytics SAS User Group Malaysia 3 rd May, 2018

2 Agenda Preface Machine Learning vs Deep Learning Create/train/score Machine Learning model using SAS VDMML Deep Learning What? Use cases How? Image classification Basic CNN architecture Layer explanation (convolution/pooling/fully connected) Deploy Create/train/score/deploy Deep Learning model using Jupyter Notebook

3 Preface

4 Machine Learning vs Deep Learning Machine Learning For structured data: Information with high degree of organization. Deep Learning For unstructured data such as: Strings of texts, images and sounds. Most experts agree that structured data accounts for about 20% of the data that is out there.

5 SAS Platform Visual Data Mining & Machine Learning (VDMML) Interactive programming in a web-based development environment. Highly scalable, distributed in-memory analytical processing. Model development with modern machine learning algorithms. Random forests, gradient boosting, neural networks, support vector machines, factorization machines, Bayesian networks. Automatic intelligent tuning. Analytical data preparation. Data exploration, feature binning and dimension reduction. Integrated text analytics. Model assessment. Model scoring.

6 SAS Platform Discovery - Basic

7 SAS Platform Discovery - Advanced

8 What? Vision Analytics Use Cases in Oil & Gas

9 Health, Safety & Environment Oil & Gas

10 Health, Safety & Environment Oil & Gas

11 Workplace Safety Leveraging on cameras placed around workplace Identify employee by name and authorization. Where is the power drill? Is the employee certified to handle a power drill? Is the power drill stored safely at the original location? Is employee equipped with the necessary PPE? Monitor if employee is trespassing. Early warning for incidents and hazardous situations. Monitor integrity of assets.

12 Workplace safety: Deep Learning on Sound

13 Vision Analytics Use Cases Oil & Gas Upstream Analyze seabed for oil seeps, which may serve as an indicator of hydrocarbon presence and for the purpose of protecting the ecosystem. Midstream Monitor hard-to-reach pipelines using bots for early signs which indicates the need for maintenance or replacements. Downstream Smarter retail experience for increased customer satisfaction.

14 Workplace Safety Example: Hard hat image detection

15 Image Classification For us, identifying whether someone is wearing a hard hat or not is effortless How do we create and train a machine to do the same???

16 How?

17 Image Classification using Deep Learning Convolutional Neural Network (CNN) Convolutional neural network to analyse images. Why? Powerful (analyse and classify images very well) Efficient (less parameters than previous methods) Source: CNN takes image (volume of pixels of varying values) and outputs probability

18 Types of CNN Architecture LeNet-5 VGG16 AlexNet Etc

19 Example CNN architecture Convolution Layer Source:

20 Convolution Layer Enables parameter sharing in a CNN

21 Convolution Layer How is it done? SWAT.deeplearn.addlayer(layer={ type : convo, nfilters :1, height :3, width :3, stride :1 } Source:

22 Feature Detector & Feature Maps Video example: Source:

23 Example CNN architecture Pooling Layer Source:

24 Pooling Layer Reduces the number of trainable parameters in a CNN

25 Pooling Layer How is it done? SWAT.deeplearn.addlayer(layer={ type : pool, height :2, width :2, stride :2, pool : MAX } Source: Video example:

26 Example CNN architecture Fully Connected Layer Source:

27 Fully Connected Layer Enables high-level reasoning Neurons in a fully connected layer have connections to all activations in the previous layer(s), as seen in regular neural networks. Information flows through a neural network in 2 ways: Normal (Feedforward) Learning (Backpropogation) AruvnKk

28 Deployment SAS VDMML (Model studio) Operational Training SAS ESP (Event streaming)

29 Creating/training/scoring/deploying a CNN Using deeplearn action sets in Jupyter Notebook on SAS Viya 3.3

30 Useful Links What s New In SAS Deep Learning (Documentation) 7ywfs6e4kn160kru9w97fyz.htm&docsetVersion=8.2&locale=en Understanding Convolutional Neural Networks Convolutional-Neural-Networks/ CS231n Convolutional Neural Networks for Visual Recognition