When it comes to creating unparalleled experiences for retail shoppers, core platform capabilities in Vision Computing and Machine Learning (ML) are at the forefront. They derive insights from structured data and unstructured content like videos, images and documents. Retailers can gain innumerable insights from this data that can be shaped into end-to-end, ubiquitous experiences, including acquiring, processing, analyzing and understanding content. Vision Computing also helps in the extraction of high-dimensional data to produce numerical or symbolic information – for example, information in the form of decisions. The ML platform enables niche developers to focus on writing efficient ML models and train, deploy as well as manage these models with minimal effort.
Vision computing as a capability is composed of a suite of products, built on an open-source stack guided by distributed micro-service architecture and cloud native:
You can build a highly scalable and reliable computer vision platform for the enterprise, by leveraging all vision IoT devices in the ecosystem to derive insights and solve for inventory management, shortage, compliance, safety and many other retail operations. This in turn can easily be extended to other retail domains like merchandizing, product design and management and supply chain.
One such example would be a platform that leverages Artificial Intelligence (AI) and ML to count the number of customers at a store, using the cameras at store entrances and exits. The solution is effective in helping stores manage crowds and queues during major product launches.
Imagine custom-building a camera and building robotics solutions on top of it for specific vision-based business use cases – like monitoring shelves remotely and generating replenishment alerts when a shelf space goes empty.
Building a solution of this complexity and scale requires inter-disciplinary collaboration – integrating with enterprise platforms like Vision, ML, IoT, Cloud environments and store facing apps. It also has the potential to scale out to other IoT devices, robotics, AR (Augmented Reality) and wearables.
The idea behind the platform is to create retail-specific vision catalog and data that can manage data from all distributed devices across thousands of sites. The pipeline can span data collection, management, curation/annotation and democratization. This becomes the foundation for any enterprise-scale Vision Computing product, service or solution.
While Vision Computing has a great start in creating future-focused technology, the journey is just beginning. There’s more to go as enterprises learn to leverage these capabilities and run them at a large scale – especially with a dynamically changing business landscape and advancements in AI as well as related technologies.