Helped build training data for AI/ML models
For a former Fortune 500 company, a leading supplier of semiconductor-based solutions
About the Customer
A former Fortune 500 company based in Arizona, USA, the client is a leading supplier of semiconductor-based solutions, offering a comprehensive portfolio of energy efficient power management innovations, empowering customers to reduce global energy use. They operate a network of manufacturing facilities and responsive, reliable, world-class supply chain and quality programs. Listed among the Top 20 semiconductor sales leaders, their products include power and signal management, logic, discrete, and custom devices for automotive, and power applications.
High-resolution images have helped extend possibilities for imaging technology and mandated a new breed of digital image sensors to support machine learning models to develop intelligent systems. The client required training data for the machine learning models to enable advanced systems, including the use of Artificial Intelligence to detect features, objects and pattern recognition for industrial automation and navigation. They were also keen to use Computer Vision to analyse and process images and videos to increase performance in low light conditions as well and improve the efficiency of their AI Platform for automatic threat detection
NextWealth’s Computer Vision capabilities were used to build training data for their AI/ML models. Models were trained with image annotations using bounding box and polygon annotations to identify, prevent and mitigate threats.
The client was able to develop a prototype image sensor inbuilt with a Machine Learning algorithm was trained to decrease false positives, thus increasing the speed at which threats could be detected. This led to dramatically shrink the window of compromise for a security system. The success of AI assisted Security Systems reiterate the fact that when humans collaborate with machines on cybersecurity, Machine Learning and Computer Vision can help them a great deal.