Deep InSight: Deep Extraction of Driver State from Naturalistic Driving Dataset
Project Information
Project ID:
FHWA-PROJ-19-0010
Project Abstract:
Computer vision algorithms have been improved significantly to provide accurate detection, tracking, and recognition of objects. Available public datasets are crucial for extracting key features thanks to ever-developing machine learning and deep learning algorithms in general. Since cameras are becoming vastly available to be employed on vehicles to extract useful information for both autonomous driving and intelligent driver assistance, the researchers are developing intelligent driver state estimation algorithms based on state-of-the-art detection and recognition using computer vision.
One of the main drawbacks for naturalistic driving data is having low-resolution and noisy video data that limits the overall accuracy when tested with the models trained on clear images. The researchers will use a comprehensive artificial intelligence (AI) platform for data management, modeling, and enhanced annotations; video quality enhancement using deep models; face detection at acute angles; and recurring network-based driver state estimation.
Project Status:
Active
Project Funding Amount (Contract Award Amount):
$1,005,125.00
Start Date:
End Date:
Public Access Plan:
No
FHWA Program Area:
- Exploratory Advanced Research
Fiscal Year / AMRP Program / AMRP Activity:
- FY 2002-2022 / Exploratory Advanced Research
AMRP = Annual Modal Research Plan
Contact Information
First Name:
Ana Maria
Last Name:
Eigen
Phone Number:
(202) 493-3168
Email Address:
Management Organizations
Sponsor Organization
Organization Role:
Sponsor Organization
Organization:
Federal Highway Administration: Office of Research, Development and Technology - Office of Safety and Operations Research and Development