From 360 photography to photorealistic holographic objects, Fyusion develops products powered by 3D computer vision and machine learning. We present a robust deep learning solution for capturing and rendering novel views of complex real world scenes for virtual exploration. Using just a regular smartphone, and with a video capture or a grid-like set of images as input, our view synthesis algorithms expand each sampled view into a local light field via a multiplane image (MPI) scene representation, then render novel views by blending adjacent local light fields. We believe that everyone should have the ability to capture, view and share 3D/lightfield images, using the smartphones that are in the market today. In this talk, we delve into the details from capture to rendering, steps in between as well as what we believe the future of digitizing the world looks like.
Pantelis Kalogiros, SVP of Web and Co-Founder, Fyusion
Pantelis is responsible for Fyusion’s cloud and web services, and is the author of Fyusion’s patented 3D viewer technology which is available on over 200 web platforms. Pantelis has 10 years of professional research and engineering experience with web, cloud and mobile technologies, specializing in optimized, fast, scalable architectures and code. He is experienced in application research and development, database design and optimization, 3D Graphics, Computer Vision, Machine Learning and UI development. Pantelis has authored the book “Practical Guide to WebAssembly” and is passionate about research, web standards and open source.