Skip to content
View eliottcrancee's full-sized avatar

Highlights

  • Pro
Block or Report

Block or report eliottcrancee

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
eliottcrancee/README.md

Welcome to Eliott Crancée's GitHub Page! 👋

AI Expert | Engineer Student | Centrale Méditerranée

💼 Experiences

  • AI Developer Intern, Infocosme, Apr 2024
  • Research Internship, Laboratory of Computer Science and Systems, Jan 2024
  • Research Internship, Laboratory of Mechanics and Acoustics, Jun 2023 - Jul 2023

🎓 Education

  • Master in AI, Aix-Marseille University, 2023 - 2024
  • Master's level engineering degree, Centrale Méditerranée, 2021 - 2024
  • Preparatory class, High school La Martinière Monplaisir, 2019 - 2021

📧 Contact

Feel free to explore my projects and connect with me on LinkedIn! Let's collaborate and innovate together! 🌟🤝

Projects

ParoleNet

ParoleNet is a multimodal model for predicting turn-taking in conversations. Its primary objective is to determine, at the end of a given sentence, whether the current speaker will continue speaking or yield the floor to their interlocutor. It takes as input the last 20 words of the sentence and the final two seconds of the audio recording, performing classification into two turn-taking classes.

🔍 Key Highlights:

  • Trained on a specific French-language dataset containing 16,400 sentences.
  • Achieved promising results surpassing random initialization, indicating the model's ability to anticipate turn-taking.
  • Statistical analysis suggests the model understands the relationship between input data and turn-taking prediction, rather than relying solely on frequency-based predictions.
  • Demonstrates significant learning, validating the effectiveness of our approach.

🔗 Project Link: ParoleNet

Pinned

  1. ParoleNet ParoleNet Public

    Utilizing a multimodal architecture to predict the appropriate speaker turn in a dialogue.

    Python