You don't know what "dropout" is: creating cutting-edge technologies on a stream [ru]
Most of the basic techniques in ML, including dropout, are used by 9 out of 10 professionals without understanding how they work.
And the creation of breakthrough technologies requires a deep understanding of the basic principles.
- What ML practice in Reface is?
- Why doesn't everyone "think and understand what you're doing"?
- How to put the production of new approaches on the stream?
- VP of Engineering at Reface.
- He's responsible for product and technology development, implementation of new ML functions, and responsible use of synthetic media.
- Tech-professional with 10 years of experience, 6 years in ML.
- Alexey is a physicist. He started studying machine learning and data science on his own 5 years ago. Prior to that, he worked as a development engineer for 6 years.In 2013 he moved to the Netherlands, where he still lives.
- He independently built a data science department at PVH Europe from scratch.
- In 2020, he published his own development of an online fitting room based on full-body-swap technology Beyond Belief.