If you're new to deep learning, this workshop is for you. Learn how to build and deploy computer vision models using the AWS DeepLens deep learning-enabled video camera. Also learn to build a machine learning application and a model from scratch using Amazon SageMaker. Finally, learn to extend that model to Amazon SageMaker to build an end-to-end AI application.

2. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Get Started with Deep Learning and
Computer Vision Using AWS DeepLens
Jyothi Nookula
Sr. Product Manager, Amazon
DeepLens
Amazon Web Services – AI
A I M 3 1 6
Kashif Imran
Sr. Solutions Architect, Amazon
Rekognition
Amazon Web Services – AI

3. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
© 2018, Amazon Web Services, Inc. or its Affiliates. All rights reserved.
Artificial intelligence at Amazon
© 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

4. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
AWS DeepLens
is not a
video camera …
It’s the
world’s first
deep learning
enabled
developer kit

5. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Platform services
Application services
Frameworks and interfaces
Caffe2 CNTK MXNet PyTorch TensorFlow Chainer Keras Gluon
AWS deep learning AMIs
Amazon SageMaker
Amazon
Rekognition
Amazon
Transcribe
Amazon
Translate
Amazon
Polly
Amazon
Comprehend
Amazon Lex
AWS
DeepLens
Education
The Amazon Machine Learning stack

6. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Get started with sample projects
Artistic style
transfer
Object
detection
Face detection/
recognition
Hot dog/
not hot dog
Cat or dogActivity
detection
Add custom functionality
or
Create your own project

7. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
2. Deploying a model
to AWS DeepLens
1. Machine learning
overview
4. Extending a
project
Today We Will Cover
3. Training a model
in Amazon
SageMaker

8. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

9. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Model training Inference
Overview of deep learning
Data

10. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Data
Annotate Preprocess Data split

11. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Model training
• Define model architecture
• Input the annotated and cleaned data into the model
• Multiple iterations (epochs) to train the model
• Validate with held back dataset
Large, annotated
dataset
Training set
Validation set
Training
Validate

12. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Inference
It’s where the magic happens!
1. Preprocess the new data or image just like a training set
2. Feed image back to the trained model to get a predicted output

13. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

14. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
15 minutes
1. Find the instruction manual here:
https://bit.ly/2DKpHjO
2. Select: Lab 1
Self-Paced Lab Instructions

15. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
AWS DeepLens Specifications
• Intel Atom Processor
• Gen9 graphics
• Ubuntu OS- 16.04 LTS
• 100 GFLOPS performance
• Dual band Wi-Fi
• 8-GB RAM
• 16-GB storage (eMMC)
• 32-GB SD card
• 4 MP camera with MJPEG
• H.264 encoding at 1080p resolution
• 2 USB ports
• Micro HDMI
• Audio out
• AWS Greengrass preconfigured
• Intel clDNN Optimized for MXNet

16. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
AWS DeepLens architecture
Video out
Data out
I n f e r e n c e
D e p l o y p r o j e c t s
Manage device
Security
Console project
management
Intel: Model Optimizer
cIDNN and
driver
AWS Cloud
MQTT
protocol
Lambda
function

17. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

18. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Amazon SageMaker
Fully managed
hosting with auto-
scaling
One-click
deployment
Pre-built
notebooks for
common
problems
Built-in, high-
performance
algorithms
Hyperparameter
optimization
Build Train Deploy
One-click
training

19. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
40 minutes
1. Find the instruction manual here:
https://bit.ly/2DKpHjO
2. Select: Lab 2
Self-paced lab instructions

20. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

21. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Amazon Rekognition:
Deep-Learning-Based Image and Video Analysis

22. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Rekognition Features
Faces Celebrities Labels
Moderation ScenesActivities Paths
Text

23. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Rekognition benefits
Low cost
Your data
is your ownFully Managed
Rapid integrationState of the
art capabilities
Continuous
improvement

24. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Media discovery
Customer
engagement Industrial
… and more
Safe content Public safety Mapping
Amazon Rekognition Customers

25. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Media discovery
Customer
engagement Industrial
…and more
Safe content Public safety Mapping
Focus of Today’s Workshop

26. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

27. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.

28. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Inference
Lambda
Amazon S3 bucket Recognize objects and
notify using IoT
Worker safety with AWS DeepLens and Amazon Rekognition
Compliance log

29. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
DetectLabels
{
"Image": {
"Bytes": blob,
"S3Object": {
"Bucket": "string",
"Name": "string",
"Version": "string"
}
},
"MaxLabels": number,
"MinConfidence": number
}
{
"Labels": [
{
"Confidence": number,
"Name": "string”
}
],
"OrientationCorrection": "string"
}
Image API – Request and Response

30. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
GetLabelDetection
StartLabelDetection
{
“ClientRequestToken": "string",
"JobTag": "string",
"MinConfidence": number,
"NotificationChannel": {
"RoleArn": "string",
"SNSTopicArn": "string”
},
"Video": {
"S3Object": {
"Bucket": "string",
"Name": "string",
"Version": "string”
}
}
}
{
"JobStatus": string,
"StatusMessage": string,
"VideoMetadata": {
"Format": string,
"Codec": string,
"DurationMillis": number,
"FrameRate": float,
"FrameWidth": number,
"FrameHeight": number
},
"NextToken": string,
"Labels": [
{
"Timestamp": number,
"Label":
{
"Name": string,
"Confidence": float
}
}
],
...
JobId
Video API – Request and Response

31. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
40 minutes
1. Find the instruction manual here:
https://bit.ly/2DKpHjO
2. Select: Lab 3
Self-paced lab instructions

32. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Now, let’s see it in action.

33. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Questions?

34. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.
Thank you!

35. © 2018, Amazon Web Services, Inc. or its affiliates. All rights reserved.