customVisionModelling.md

Object Detection Models using Custom Vision

Steps

1. Create resources on Azure Portal
2. Data preparation 3 Upload images into Azure ML compute instance
3. Extract tags by converting COCO to PASCAL VOC format

Step 1 - Create resources on Azure Portal

Step 2 - Upload images into Azure ML compute instance

Step 3 - Create a Custom Vision project

Step - Extract tags by converting COCO to PASCAL VOC format

Step - Resize images and normalize tag coordinates

Useful Resources

• Python SDK Docs: https://docs.microsoft.com/en-us/python/api/azure-cognitiveservices-vision-customvision/?view=azure-python
• Python Custom Vision Client Lib: https://pypi.org/project/azure-cognitiveservices-vision-customvision/
• Azure SDK for Python: https://github.com/Azure/azure-sdk-for-python/