Neural network-based chess engine capable of natural language commentary
-
Updated
Nov 15, 2022 - C++
Neural network-based chess engine capable of natural language commentary
Automatic Baseball Commentary Generation Using Deep Learning (SAC 2020)
An automated LLM-based Go game commentary generator
Github Repository for CommBox-A system aimed at identifying cricket shots and generating automatic commentary using sensors and machine learning
End-to-end implementation of our research paper "Bridging Chess Engines and Language Models: A Novel Framework for Generating Human-Like Chess Commentary"
Experimental LSTM + Multi-Head Attention Model + LLM decoding strategies for Football Commentary Generation
Euro 2024: Data-Driven Momentum Analysis & AI Commentary DS-AI project where data leads methodology. Statistics, pattern discovery and autocorrelations revealed 3-minute momentum windows and final 15-minute criticality. An AI agent evaluates relevance and forwards selected data to a fine-tuned LLM for ESPN-style commentary with ARIMAX predictions.
AI-powered football video analysis with player tracking, commentary generation and highlight detection
Our model generates descriptive captions for the given input image related to cricket domain.
To associate your repository with the commentary-generation topic, visit your repo's landing page and select "manage topics."