Researchers from Skoltech and the AIRI have developed a new algorithm for optimal data transfer between domains using neural networks

The Artificial Intelligence Research Institute (AIRI) and Skoltech have collaborated on a novel algorithm for optimizing information sharing across disciplines using neural networks. Unlike other approaches that need coupled training datasets like input-output examples, the novel approach may be trained on separate datasets from the input and output domains. Recent advances in generative models have made this job much easier by dramatically improving the quality of generated text and images. In other words, generative models facilitate the transition from one domain to another by synthesizing data from different data. Neural Optimal transport is a novel approach that uses deep neural networks and separate datasets.