Deep Learning for Object Detection Series

Copyright 2019-2023 The MathWorks, Inc.

Introduction

This repository consists of the code files for the following videos; the dataset used for this code can be found here: https://drive.google.com/drive/u/1/folders/1bhohhPoZy03ffbM_rl8ZUPSvJ5py8rM-

1. Data Preprocessing for Deep Learning
2. Design and Train a YOLOv2 Network in MATLAB
3. Import Pretrained Deep Learning Networks into MATLAB
4. Deploy YOLOv2 to an NVIDIA Jetson
5. How to Perform Deep Learning Inference in Simulink

Set-Up

1. Clone the repository.
2. Launch Deeplearningforobjectdetection.prj to set-up the project environment.
3. Download the data from this folder: https://drive.google.com/drive/u/1/folders/1bhohhPoZy03ffbM_rl8ZUPSvJ5py8rM-
4. The first time you run this example, edit appropriately and run adjustGroundTruthPaths.m to adjust the ground truth data objects' source path for your computer. Instructions on the edits needed are in the code file.

Folder Structure

1. codeFiles folder consists of MATLAB® code files for each of the videos and a sample Simulink® model
2. Utilities folder consists of .MAT files required for uplaoding pre-trained content and helper functions

Additional Resources

Labeling data:

1. How To Label Video and Image Data For Deep Learning
2. How to Use Custom Automation Algorithms for Data Labeling

Training an ACF Detector for custom automation algorithms:

1. Code & Videos: Using Ground Truth for Object Detection

For any questions contact the authors: roboticsarena@mathworks.com