AWS Computer Vision Jump Starter Kit

author: dylatong@amazon.com

---

This repository is a collection of content to help enable engineers and data scientists to succeed on their Computer Vision projects on AWS. The repository currently includes labs for:

1. Amazon Rekognition Custom Labels: Computer Vision AutoML for Image Classification and Object Detection
2. Amazon GroundTruth: Managing Machine Learning Annotations at Scale
3. Amazon SageMaker Built-in Algorithms: A brief follow-up to the GroundTruth Lab for building a SSD Object Detector using the SageMaker built-in Algorithm.
4. Amazon SageMaker and GluonCV for Object Detection: Training and Deploying YOLOv3 on GluonCV and SageMaker.
5. Amazon SageMaker and GluonCV for Pose Estimation: Training and Deploying an Inference Pipeline for 2D Human Pose predictions.

Lab guides are shared here.

2020 Overview Presentation is shared here

Related Content:

• 2019 Image Classification and Object Detection Partner Webinar
• Image Similarity using PyTorch AWS ML Blog (references to repository)
• Jetson Nano Smart Cam Repository

Sample Workshop Packages

Examples of modules that you can use together to tailor a workshop for your audience.

I. Object Detection Workshop Package (est. 4 hours)

Below is content you can package up into a Object Detection workshop for SageMaker. You can put together a 4-5 hour agenda with this content.

1. AWS CV Introduction Presentation

2. Lab1: Ground Truth:

   • Learn to create and manage a quality data set at scale using SageMaker GroundTruth.
   • Manage annotation workforces: private, public (Mechanical Turk), and 3rd party vendors.
   • Create a labeling job (for Object Detection)

3. Lab2: SageMaker Algorithms- Object Detection:

   • Learn to build a custom object detection (Single-shot Detection) from the training data you created in Lab1 without having to write code.
   • Learn about hyper-parameter tuning automation.

4. Lab3: Bring Your Own Script- Object Detection:

   • Learn how to bring your own script from a deep learning framework.
   • In the lab we’ll bring a GluonCV script to train an object detection model (YOLOv3 on mobileNet).
   • Learn how to programmatically launch a hyperparameter tuning job, SageMaker local training as well as perform incremental training.
   • Learn how to deploy a real-time endpoint for inference.

II. Complete AWS Computer Vision Workshop (est. 16 hours)

1. AWS CV Introduction Presentation
2. Rekognition Custom Labels Lab
   • Provides a first hands-on experience with Rekognition Custom Labels through a logo detection use case.
   • Learn the iterative process and get a sense of how to extend the learnings to deliver a production-ready deployment.
3. [Object Detection Series]:

   • Learn the end-to-end experimentation process for CV projects on Amazon SageMaker

   • Learn the different ways to approach CV problems on Amazon SageMaker and acquire an understanding of the pros and cons.

   • Lab1: Ground Truth:

     • Learn to create and manage a quality data set at scale using SageMaker GroundTruth.
     • Manage annotation workforces: private, public (Mechanical Turk), and 3rd party vendors.
     • Create a labeling job (for Object Detection)

   • Lab2: SageMaker Algorithms- Object Detection:

     • Learn to build a custom object detection (Single-shot Detection) from the training data you created in Lab1 without having to write code.
     • Learn about hyper-parameter tuning automation.

   • Lab3: Bring Your Own Script- Object Detection:

     • Learn how to bring your own script from a deep learning framework.
     • In the lab we’ll bring a GluonCV script to train an object detection model (YOLOv3 on mobileNet).
     • Learn how to programmatically launch a hyperparameter tuning job, SageMaker local training as well as perform incremental training.
     • Learn how to deploy a real-time endpoint for inference.

   • Lab4: BYOS and Inference Pipelines- Pose Estimation:

     • Learn how to implement an inference pipeline of multiple Computer Vision models.
     • Learn how to build and deploy a pose estimation model.
4. AWS CV@Edge online serices
   • This is a online series with hands-on labs to teach you about CV@edge and associated tools available to you on AWS.