LIA-RAG: a system based on graphs and divergence of probabilities applied to Speech-To-Text Summarization

TL;DR

This paper presents LIA-RAG, a novel speech-to-text summarization system that utilizes probabilistic divergence and graph-based methods, demonstrating promising results on a multilingual French speech corpus.

Contribution

The paper introduces a new algorithm combining probability divergences and graph techniques for speech-to-text summarization, tailored for noisy speech data.

Findings

Encouraging results on CCCS Multiling 2015 French corpus

Effective handling of noisy speech input

Potential for improved speech summarization performance

Abstract

This paper aims to introduces a new algorithm for automatic speech-to-text summarization based on statistical divergences of probabilities and graphs. The input is a text from speech conversations with noise, and the output a compact text summary. Our results, on the pilot task CCCS Multiling 2015 French corpus are very encouraging