How to Use Facebook's DETR Object Detection Algorithm in Python - Full Tutorial

Overview

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!

Grab it

Follow along with this comprehensive tutorial video on implementing Facebook's DETR (DEtection TRansformer) object detection algorithm using Python and PyTorch. Learn how to load the pre-trained model from TorchHub, prepare images for input, process model outputs, and visualize detection results with bounding boxes. Explore various image types including abstract art and optical illusions to test the model's capabilities. Discover techniques for adjusting prediction thresholds and handling challenging scenarios. Gain practical insights into advanced computer vision techniques through hands-on coding examples and real-world applications.

Syllabus

- Intro
- TorchHub Model
- Getting an Image
- Image to PyTorch Tensor
- Handling Model Output
- Draw Bounding Boxes
- The Dress
- Rorschach Ink Blots
- Forcing More Predictions
- Jackson Pollock Images
- Elephant Herds