This course teaches learners how to train Object Detection Transformers on a custom dataset using DETR as an example. The learning outcomes include setting up a Python environment, downloading custom datasets, building PyTorch COCO Detection datasets, visualizing dataset entries, creating data loaders, building a PyTorch Lightning DETR module, training the model, performing inference, and evaluating the model. The course focuses on hands-on practical skills in Python programming, PyTorch, and object detection techniques. The teaching method involves step-by-step demonstrations and practical exercises. This course is intended for individuals interested in computer vision, deep learning, and object detection, particularly those with some prior knowledge of Python and machine learning concepts.
How to Train Object Detection Transformer on Custom Dataset
Overview
Syllabus
Introduction
Setting up the Python environment
DETR model inference on example images
Download custom dataset from Roboflow Universe
Building custom PyTorch COCO Detection datasets
Visualising COCO datasets entry
Building custom PyTorch Data Loaders
Building custom PyTorch Lightning DETR Module
Training DETR on custom dataset
Custom DETR model inference
Evaluating custom DETR model
Outro
Taught by
Roboflow
Related Courses
Related articles
Reviews
4.0 rating, based on 1 Class Central review
-
Avnish PanwarA great introduction on how to use DETR on custom dataset.
But could have given a more thorough description of the internal workings or components of the model.
