Driver Behavior Recognition via Interwoven Deep Convolutional Neural Nets with Multi-stream Inputs

11/22/2018
by   Chaoyun Zhang, et al.
0

Recognizing driver behaviors is becoming vital for in-vehicle systems that seek to reduce the incidence of car accidents rooted in cognitive distraction. In this paper, we harness the exceptional feature extraction abilities of deep learning and propose a dedicated Interwoven Deep Convolutional Neural Network (InterCNN) architecture to tackle the accurate classification of driver behaviors in real-time. The proposed solution exploits information from multi-stream inputs, i.e., in-vehicle cameras with different fields of view and optical flows computed based on recorded images, and merges through multiple fusion layers abstract features that it extracts. This builds a tight ensembling system, which significantly improves the robustness of the model. We further introduce a temporal voting scheme based on historical inference instances, in order to enhance accuracy. Experiments conducted with a real world dataset that we collect in a mock-up car environment demonstrate that the proposed InterCNN with MobileNet convolutional blocks can classify 9 different behaviors with 73.97 accuracy. Our architecture is highly computationally efficient, as it performs inferences within 15ms, which satisfies the real-time constraints of intelligent cars. In addition, our InterCNN is robust to lossy input, as the classification remains accurate when two input streams are occluded.

READ FULL TEXT

page 1

page 3

page 4

page 6

page 7

page 9

page 10

research
07/03/2020

Real-Time Monitoring and Driver Feedback to Promote Fuel Efficient Driving

Improving the fuel efficiency of vehicles is imperative to reduce costs ...
research
12/10/2018

An Intelligent Safety System for Human-Centered Semi-Autonomous Vehicles

Nowadays, automobile manufacturers make efforts to develop ways to make ...
research
06/07/2017

Driver Action Prediction Using Deep (Bidirectional) Recurrent Neural Network

Advanced driver assistance systems (ADAS) can be significantly improved ...
research
11/13/2019

Vehicle Re-identification: exploring feature fusion using multi-stream convolutional networks

This work addresses the problem of vehicle re-identification through a n...
research
10/09/2017

Driving Behavior Analysis through CAN Bus Data in an Uncontrolled Environment

Cars can nowadays record several thousands of signals through the CAN bu...
research
10/13/2020

Robust Two-Stream Multi-Feature Network for Driver Drowsiness Detection

Drowsiness driving is a major cause of traffic accidents and thus numero...
research
02/04/2019

A Two-Stream Siamese Neural Network for Vehicle Re-Identification by Using Non-Overlapping Cameras

We describe in this paper a novel Two-Stream Siamese Neural Network for ...

Please sign up or login with your details

Forgot password? Click here to reset