Abstract
In this talk I will talk about Pigo the only existing “pure” Go face detection and facial landmark points detection library which does not require any third party module and is platform agnostic. I will discuss about it’s integration into other programming languages and porting it to Webassembly. Year-by-year the Go language is evolving as it increases the projects and libraries built around it, but still the Go ecosystem is still well behind its competitors in domains like computer vision, machine learning and data analysis. The talk will cover the following topics: - The majority of the existing small number of Go face detection libraries are actually bindings to some C or C++ based libraries. - Why it's important to have a single portable binary file and how to escape from the burden of compiling a huge monolithic library like OpenCV. - Possible use cases and integrations - Pupils/eyes localization - Facial landmark points detection - Running the library in different environments and programming languages as a shared object (.so) - Porting to Webassembly - I will show some hot real time demos.