Computer Vision (CV) models are used to solve a variety of problems, to name a few –
Identifying defects in parts given an image, Self-driving cars, Identify intruders in a given region, Generative AI to generate image given text, Matching faces between face in id card & face in selfie for KYC, Solving for OCR Further, CV models are deployed in various environments: Raspberry devices, Traditional k8s environment, On browser. Finally, some CV models are queried in real-time, near-real time and batch mode
While building the above solutions takes time, building sufficient guardrails to operationalize the models takes even more time.
This session will walk the user through modularizing code in such a way that they are able to build low code solution & operationalize within their organization at least 3x faster. Here are the key learning objectives from this session:
- Understand the best practices of modularizing code that can be used across solutions
- Understand the best practices of deployment for the variety of deployment as well as scale
- Learn about the guardrails to implement while exposing CV models to external usage
- Understand the best practices of leveraging cloud solutions”
