Deep learning provides solutions for many hard problems in game development, such as AI, animation, speech recognition and synthesis. However, many engines are written in C++ and this won't change in the forseeable future. This talk gives a glimpse at integration of deep learning models, trained in python into an engine written in C++. The framework used in this talk is pytorch/libtorch, the game engine we use is Godot.
About the speaker:
Georgy Derevyanko, PostDoc at Concordia/MILA, working on applications of deep learning in molecular biology (protein folding and interaction). He did his PhD at Universite Joseph Fourier (Grenoble, France) in computational predictions of protein interactions.