Description

HSC's Core Image and Video Analytics (CIVA) engine is a generalized classification (what) and segmentation (where) engine that works on both videos and images. It is coupled with HSC's big data framework accelerator to provide an easy to use system that offers the following benefits:

• Support analysis of still images, recorded  videos or streaming video
• Offers both scale-up (large systems) and scale-down (embedded boards) configurations
• Ability to chain/combine multiple machine learning models to achieve the end goal
• Fully API based, easy to integrate with your end system

Customers involved/interested in the general space of image/video analytics will know that there are many different models/approaches to solve different problems. Depending on the dataset and needs, significant knowledge is required to know which works best and the right approach - should we discard our current approach and look for a new model, or should we tune our current model?  That is exactly where CIVA helps. We do the hard work for you.

Our General Approach to Solving Diverse Problems

Features

• API based
• Already supports many inbuilt models like: Person detection and counting, vehicle counting, traffic violations, face recognition
• Allows detection of custom objects and situations
• Can be used to analyze static images or videos
• Also offers advanced person tracking features 

Use Cases:

• Deployed in a fleet security solution where our solution is used to track drivers for specific infractions, like driving without a seat belt, jumping red lights
• Deployed in a consumer electronics OEM to detect physically defective manufacturing parts compared to good quality parts
• Deployed in a retail chain to perform people counting and demographic analytics