Requirements: DNJ Indiego was approached by the client to provide a solution that can train custom data. The client wanted the solution to train and create models for different industry sectors.

Solution: We developed a framework (Graphical User Interface application) based on Deep Learning methods and TensorFlow that can be used to annotate and train custom data to create the Machine Learning model. The developed framework had the following features:

• Annotation tool: To label the image
• Train: To train the images along with the annotated data
• Predict: For testing the images to check whether the objects are detected properly

The custom models were used to build different applications for:

• Industrial Automation
• Vehicle Identification for Insurance
• Retail store to identify the packets in the shelf