Deep Learning

Dragonfly's Deep Learning applications let you advance your image processing results to new levels by harnessing the power of convolutional neural networks (CNNs). CNNs mimic human intelligence to distinguish anomalies and features while tolerating natural variations in complex patterns and offer an advantage over traditional machine learning approaches, which struggle to distinguish variability and deviation between very visually similar phases.

Dragonfly's deep learning-based algorithms are optimized for real-world image analysis, requiring smaller datasets, and shorter training and validation periods, while delivering fast, accurate results for denoising, super resolution, and semantic segmentation.

In This Section

Videos About Deep Learning

Key Features

Workflow and Data Preparation

Deep Learning Architectures

Deep Learning Tool Interface

Training Deep Models for Denoising

Training Deep Models for Super Resolution

Training Deep Models for Semantic Segmentation

Previewing Training Results

Applying Deep Models

Evaluating and Comparing Deep Models

Resources

Additional information about deep learning, videos, and links to external references are available on the Deep Learning webpage (www.theobjects.com/dragonfly/deep-learning.html).