Skip to content

Overview

Project Overview

CHAIS: Crossmodal Human-AI Collaboration for Selective Hearing Assistance

Project Proposal: CHAIS.pdf

The CHAIS project is about developing bio-inspired crossmodal Speech Enhancement algorithms to address the Cocktail Party Problem. This should be showcased in prototypical smart Glasses that support humans during selective hearing tasks. The speech enhancement should utilize mechanisms informed by cognitive speech processing in humans to (1) potentially improve real-time speech enhancement and (2) gain further insights into how these mechanisms improve speech comprehension.

The project is a cooperation between three Project Partners: 1. Human-Computer Interaction Develop human usable assistive glasses, based on user needs and feedback. 2. Signal Processing Improve AI-based real-time speech enhancement for hearing assistance in humans. 3. Neuronal Information Processing & Computational Neuroscience Apply and gain insights into cognitive mechanisms in speech processing.

Table of Contents

  • Cocktail Party Problem Problem definition and insights.
  • Glasses Project outcome: Glasses that enhance speaker in noisy, multi-speaker environments.
  • Existing Technology Overview on the current state of commercial and academic speech enhancement solution in multi-speaker scenarios.
  • Roadmap Overview of the current project roadmap.
  • Resources Available resources for the project.
  • Project Partners Expectations by project participants and potential conflicts.
  • Training & Evaluation Datasets and virtual testing environment for quick iterative development.
  • Papers to Read Relevant papers for research projects that have not been reviewed yet.