3 Day Object Detection Challenge
Build your own object detection model from start to finish and learn foundational computer vision skills.
Why joining this Challenge
Practicing Object Detection
Applying YOLO +
COCO
Presenting Results in Video Output
What is Object Detection?
Object detection is a computer technology related to computer vision and image processing that deals with detecting instances of semantic objects of a certain class (such as humans, buildings, or cars) in digital images and videos with a bounding box.
Object detection algorithms typically leverage machine learning, deep learning and convolutional neural networks (CNN) to produce meaningful features from input images and classify them into various object classes.
Object Detection Use Cases
Autonomous Driving
Medical Image Analysis
Video Surveillance
Contactless Checkout
Defect Detection
Object Counting
3 Days of Learning
Challenge Day 1
The first day of the Challenge is about getting started. You will learn more about the different versions of YOLO and set up the data & code repository.
- Overview of the Challenge (objective, accomplishment, and prerequisites)
- Understanding YOLO (“You Only Look Once” ) and its different versions
- Getting familiar with COCO dataset and its labels and format
- Reviewing the PyTorch framework that will be used for the Challenge
- Setting up code repository and data (importing libraries, etc.)
Challenge Day 2
The second day of the Challenge is about training the model, testing the performance of the trained model through validation, and evaluating the model.
- Getting to know all the parameters of YOLOv5 to effectively monitor the training
- Understanding the model architecture, different types of layers and their functions
- Choosing the hyperparameters according to dataset and training the model
- Analyzing results by interpreting loss variations on training and validation datasets
- Understanding evaluation metrics and model validation (precision, recall, F1 score)
Challenge Day 3
- Randomly selecting a number of images and copying them into a folder
- Performing inference on selected images, i.e. detect and identify objects
- Concatenating the frames that were inferenced and creating video output
- Reviewing the object detection process and learnings from the Challenge
- Sharing your output with your community and adding it to your portfolio
What students say
Marissa S.
Software Engineer
“I’m a software engineer and want to get more into data science. This 3 day object detection Challenge was an ideal intro to learn the entire process.”
Ray N.
Computer Vision Student
“For me, computer vision is absolutely fascinating and I was looking for a training to do my first object detection project. I found it with this Challenge!”
Howard F.
Data Scientist
“At the end, it is always about practice to improve ML skills. This 3 day object detection Challenge was perfect to learn and accomplish an actual result.”