0:00 Introduction
0:55 Setting up Anaconda, CUDA, and cuDNN
4:46 Installing TensorFlow
6:47 Preparing our Workspace and Virtual Environment Directory Structure
12:56 Preparing the Dataset
15:12 Generating Training Data
17:19 Configuring the Training Pipeline
22:20 Training the Model
26:23 Exporting the Inference Graph
29:01 Testing out the Model

Hi! Today we train a Custom Object Detector with the newly introduced Custom Object Detection API for TensorFlow 2.0 versions.

The steps mentioned follow my Github Repo Here: https://github.com/armaanpriyadarshan/Training-a-Custom-TensorFlow-2.x-Object-Detector

Please raise an issue in the Github Repository if you are having any errors. I am currently having trouble with comment visibility on YouTube.

You are welcome to contact me directly through Discord. I’ve seen some issues with comments not showing up, so I’ll just put my Discord Username here. Feel free to send me a DM. Troy and Abed#6396

Last tested on 8/11/2020 with TensorFlow GPU 2.3.0 on Windows 10

If you find anything cool or have any issues, feel free to let me know in the comments! If you enjoyed don’t forget to like and subscribe!

source