Project Overview

• Development of a real-time voice communication system integrating Speech-to-Text (STT), Natural Language Processing (NLP), and Text-to-Speech (TTS).
• The project aims to create a robust and low-latency communication system that supports multiple languages (French, Arabic, English).

Technical Tasks

• Implement a local chain: VAD → STT → NLP/LLM → TTS.
• Develop a dialogue manager to handle context, errors, and barge-in scenarios.
• Integrate the system with ROS2 for effective communication between components.
• Optimize audio processing and latency on Raspberry Pi 5.
• Conduct real-world testing in noisy environments with multilingual support.

Key Skills Required

• Proficiency in Natural Language Processing (NLP) and Large Language Models (LLM).
• Experience with Speech-to-Text (STT) and Text-to-Speech (TTS) technologies.
• Familiarity with ROS2 and Audio Digital Signal Processing (DSP).
• Skills in edge optimization techniques for improved performance.

For applications, please contact us at contact.nexor.robotics@gmail.com with the subject line: Application for PFE 2026 at Nexor Robotics.