Object Recognition and IoT
The privately-held company’s software helps F50 companies processing massive package volumes avoid costly delays and backups by monitoring thousands of package scanning stations. In addition to IoT data, each site ingests millions of camera images daily.
To add value by eliminating their customers’ laborious manual tasks, our client sought to use AI to analyze the images and classify/identify objects in the pictures.
Unable to use open-source models and train a naive algorithm by themselves given the uncommon object types, they turned to Gravitate for help.
At a fraction of the cost and time of the prior contractor, Gravitate delivered an MVP that included:
An image data pre-processing pipeline to clean up raw images
Customized deep-learning algorithms, with source code
An API with documentation
After the initial MVP, the client contracted for additional deep-learning image classification models, and training for the internal development team.