Description
The Translational Research for Injury Prevention Laboratory, under the leadership of Dr. Despina Stavrinos, has collected significant data on the communication of drivers in a simulated driving environment. This research utilizes three network simulators where six participants are outfitted with headsets and cameras to capture high-fidelity audiovisual data as they collaborate as a team to navigate through the simulation. Analyzing these communication data is vital for understanding group dynamics, enhancing safety measures, and informing the design of intelligent systems in real-world driving scenarios. This project presents a unique opportunity to leverage AI and advanced data analysis in the exploration of human communication within simulated environments.
Duration
Total project length: 175 hours
Task ideas
The primary objective of this project is to develop an AI-driven tool that can process and analyze communication data collected during driving simulations. Key features of the tool will include:
- Data Ingestion: Implement robust data pipelines for importing and preprocessing audiovisual communication data from simulation studies, ensuring scalability and data integrity.
- Data Processing: Leverage advanced Natural Language Processing (NLP) techniques, including speech recognition, named entity recognition (NER), and intent detection, to analyze spoken communication.
- Key Variable Extraction:
- Speaker Identification: Utilize speaker diarization algorithms to accurately identify and differentiate speakers within the audio stream.
- Timestamping: Implement precise timestamping mechanisms for each utterance to maintain chronological context.
- Content Analysis: Apply sentiment analysis and topic modeling to extract meaningful insights from the dialogue.
- Dynamic Reporting: Generate comprehensive communication metrics, including total counts and engagement levels by individual and group. Visualize communication dynamics over time using interactive heat maps and network graphs that illustrate speaker interactions and conversational flows.
- User Interface: Develop a responsive and intuitive user interface using modern web technologies, allowing researchers to interact seamlessly with the data and generate customizable reports.
- User-Centric Customization: Enable advanced customization options, allowing users to tailor dashboards and visualizations based on specific research queries and interests.
Expected results
- The successful project will develop a fully functional prototype of the AI-driven communication analysis tool, incorporating advanced NLP and visualization techniques. Comprehensive documentation and a user guide to facilitate adoption and usage are also expected.
Requirements
The project requires the ability to code in Python and knowledge of machine learning and natural language processing.
Project difficulty level
This project has a moderate level of difficulty.
Mentors
Please DO NOT contact mentors directly by email. Instead, please email human-ai@cern.ch with Project Title and include your CV and test results. The mentors will then get in touch with you.
Corresponding Project
Participating Organizations