Human Factor Study: Driver Behavior Recognition

Human Factor Study: Driver Behavior Recognition

DENSO is one of the biggest tier one suppliers in the automotive industry, and one of its main goals is to provide solutions to the OEMs in the Automated Driving field. DENSO plans to increase its contribution to the development of automated driving technologies by enhancing the functionalities of existing techniques and invent new solutions.

Predicting “next step” behavior and responding appropriately is the key to autonomous navigation.  One of the elements in this environment is the driver or operator of the primary vehicle. By understanding the current environment and needs of the driver we can better predict their optimal environment in the “next step”.  Will they need information? Automatic evasive maneuver? Etc.

This student team will improve our status quo driver evaluation through both Audio- and Video-based evaluation.

Project Topics include:

Audio based sentiment classification with deep learning

  1. Speech recognition (Speech to text): Conversion of spoken words to text with recurrent neural network.
  2. Voice recognition: Extract voice features and representations to a vector and use deep neural network to classify who is the speaker.

Video-based face detection and recognition with deep learning

  1. Detect, locate and track the human face in an image or video
  2. Face recognition: Identify or verify a person based on the face detected with deep neural network.
  3. Build a face recognition system with GUI

In this advanced project, all work will be done within the context of Prof. Kayvan Najarian’s research. Students will assign all project-related IP to the University of Michigan.

More Information Coming Soon

 

Students who successfully match to this project team will be required to sign the following document in January 2018:

Click here to view Student IP Agreement

No NDA necessary

How to Apply

Project Features

  • Skill level Graduate
  • Students 5-7 Students
  • Likely Majors CE, CS, DATA, ECE, EE, IOE, ISD-AUTO, ISD-GAME, MICDE, MIDAS, STATS
  • Course Substitutions Honors, MICDE, MIDAS, Data Science, ISD-Auto, ISD-Game, ISD-MFG, ECE Cognate, EECS 498, ROB 590, IOE Capstone, IOE Grad
  • IP & NDA Required? Yes
  • Summer Opportunity See Complete Description for Details
  • Artificial Intelligence and Machine Learning (3 Students)

    Python programming, opencv, basic machine learning/deep learning understanding

    • Likely Majors: EE, CE, CSE/CS-LSA, ROB
  • Programming (2 Students)

    General Programming (Python)

    • Likely Majors: CSE/CS-LSA
  • Applied Data Analysis & Statistics (1 Student)

    Numeric methods, statistics, stochastic calculus, or other statistical packages, must also have basic coding experience

    • Likely Majors: MIDAS/MICDE, DS, STATS, IOE

Faculty Mentor: Kayvan Najarian
Associate Professor, Computational Medicine and Bioinformatics, and Emergency Medicine
Dr. Kayvan Najarian serves as the director of M-CIRRC Biosignal-Image and Computational (BIC) Core program. Dr. Najarian received his Ph.D. in Electrical and Computer Engineering from University of British Columbia, Canada, M.Sc in Biomedical Engineering from Amirkabir University, Iran, and B.Sc. in Electrical Engineering from Sharif University, Iran. The focus of Dr. Kayvan Najarian’s research is on the design of signal/image processing and machine learning methods to create computer-assisted clinical decision support systems that improve patient care and reduce the costs of healthcare. Dr. Najarian’s lab also designs sensors to collect and analyze physiological signals and images. In particular, Dr. Najarian’s research focuses on creating decision support systems to manage traumatic brain injuries, traumatic pelvic/abdominal injuries and hemorrhagic shock, cardiac arrest and other critical care states. Dr. Najarian’s research has been funded by agencies such as National Science Foundation and Department of Defense. He serves as the Editor-in-Chief of Biomedical Engineering and Computational Biology and the Associate Editor of two other journals in the field of biomedical informatics. He is also a member of the editorial board of many other journals and serves as the guest editor of special issues for several journals in the field. Dr. Najarian has over 150 peer-reviewed journal and conference publications including a highly referenced textbook in the field of biomedical signal and image processing.