Introduction

Datagen(Data synthesis) is the process of generating images with accurate analysis based on sample images. The process simplifies the need for decoding complex scenarios by reading and analysing images.

Applicability

• Helps in creating images of Rare Data Types such as defects on metal surfaces.

• Enables simulation of scenarios in manufacturing and production industries that involve development and testing of AI models.

• Helps in creating datasets for medical imaging to assist in training machine learning models aimed at diagnosis and detection of diseases.

• Allows generation of datasets for training models applicable to security and surveillance systems.

How does it work?

For illustration purposes, let’s understand how Datagen(Data Synthesis) helps in synthesizing concepts on images and generating the required files for a dataset.

Prerequisites

• To synthesize and generate a dataset of images, you will need to create a project.

• To generate files based on the concepts, you need the image files for synthesis.

Datagen(Synthesis) process

The process involves four core steps:

1. Provide inputs with images: This involves marking the concept and location on images and adding image samples as background on which the concept should be synthesized.

2. Preview the image dataset: Build a preview session to generate the dataset based on the concept marked.

3. Review the synthesized dataset: Verify if the dataset is accurately displaying the defects as expected.

4. Accept and generate the final dataset of image.

You can next download the generated files and register them under a dataset to proceed with the next steps.

The following diagram depicts the overall Datagen flow in Data Explorer.