Train a Production-Level Image Classifier

Overview

Learn how to train a production-level image classifier in this comprehensive 23-minute tutorial from Wolfram. Explore the entire training workflow, from preparing datasets to implementing advanced techniques like image augmentation, transfer learning, and learning rate scheduling. Discover how to create and apply templates, conduct experiments, and set up remote training for more efficient model development. Gain insights into future directions in image classification as you master the skills needed to build robust and accurate classifiers for real-world applications.

Syllabus

Introduction
Recap
Training Dataset
Training Network Architecture
Training Workflow
Basic Workflow
Configuration File
Image Augmentation
Basic Idea
Dropout Layers
Label Smoothening
Transfer Learning
Learning Rate Schedule
Blind Training
Create Notebook
Create Template
Apply Template
Experiments
Future Directions
Remote Training