Experimental AI chessboard with computer vision, generative AI (Google Gemini or Anthropic Claude)

Hi all, I have made an Experimental AI chessboard with computer vision, generative AI (Googlel Gemini or Anthropic Claude), Web Speech API to control and communicate with voice and connected to Stockfish API for oponent moves. I put it publicaly available this week (17. 1. 2025). You use it with your smartphone which is doing all needed including the computer vision stuff to get the moves directly from the wood chess board. It is able to recognize the moves and the figures on the board. There are couple options for the setup and you can also use it without comp vision and control it with voice only. In such case you can use it directly with your own board at home. Or you can also make your own TensorFlow model for comp vision and then play it again on your chess board with your figures. Anyway it is still work in progress but already available for anybody testing if interested. You can get more details on the website of 4makers organisation or see a short video on youtube (Computer Vision - Experimental AI chessboard). Enjoy... and I'll be happy for your comments.