#100DaysofMLCode challenge involves coding and/or studying machine learning for atleast an hour everyday for the next 100 days. I am pledging today that I will be taking this challenge starting today and I will post my daily updates as a log in both github and here in my website as a blog post.
Here is the today update, I’m currently in my Week 3 Programming assignment for Coursera’s Convolutional Neural Networks course (part of Deep learning specialization). This assignment is about “Autonomous driving - Car detection”
I will be learning about object detection using the very powerful YOLO model. Many of the ideas in this notebook are described two papers: Redmon et al., 2016 and Redmon and Farhadi, 2016
The objective is to learn,
- Using object detection on a car detection dataset
- How to deal with bounding boxes
- Problem Statement
If you are working on an autonomous self-driving car, the most critical part of this project is to build a car detection system. To collect data, a mounted camera on the hood, which takes pictures for every few seconds while you drive. The dataset that was used in this assignment was obtained by drive.ai
- YOLO 2.1 Model details 2.2 Filtering with a threshold on class scores 2.3 Non-max suppression
2.4 Wrapping up the filtering
