基于深度学习的驾驶员分心驾驶行为（疲劳+危险行为）预警系统使用YOLOv5+Deepsort实现驾驶员的危险驾驶行为的预警监测

基于深度学习的驾驶员状态检测，不仅仅可以识别出疲劳驾驶，还能够识别出各种各样的状态

Keras implementation of Human Action Recognition for the data set State Farm Distracted Driver Detection (Kaggle)

10th place solution in "State Farm Distracted Driver Detection"

Solves a kaggle problem of State Farm Distracted Driver Detection

Predicting the likelihood of what the driver is doing in each of the pictures in the dataset.

Automatically classification of each driver's behavior given a dataset of 2D dashboard camera images.

StateFarm dataset used to predict the class of the distracted driver using VGG-16, RESNET50, XCEPTION and MOBILE NET models

This project aims to detect the dangerous status of driving based on the images captured by the dashboard camera using deep learning.

A video-based distracted driver detection system.

Starter project for the Kaggle State Farm Distracted Driver Detection Competition

# Intellifotainment assist” – Smart HMI for passenger cars To run the program, download all files and save them in the same directory. After that, simply run 'Master.py' in the terminal. At the moment, the program will only run in linux based systems. # The Idea Infotainment systems have come a long way since the first set of dashboards installed in cars. Through our idea, we aim to create a Human Machine interaction model that takes infotainment systems to a new level. The driver tends to get distracted from the road while performing secondary tasks such as changing the music track, locking/unlocking the door while driving etc. Our system shall enable the driver to focus only on driving. Controlling the secondary tasks will be much easier. Our product primarily comprises of 5 modules: 1) Attention and drowsiness detection: - A camera shall be present on the dashboard, in front of the driver, behind the steering wheel. Through digital image processing techniques , using hough circle algorithm and haarcascade of an eye, we shall keep track of the driver’s sight. If he or she is looking away from the road while driving for more than a specified amount of time, we shall alert the driver to focus. We shall map the head orientation and iris position to accurately identify the driver’s attention. 2) Infotainment control features using blink combination: Through a combination of blinks, the driver can turn on or of the headlights, tail lights as well as indicators. Blinking of the eyes shall be detected using ‘dlib’ features in python. This shall give extremely accurate results. 3) Voice commands to control wipers, car lock, music system and windows A simple, yet extremely useful idea that would make the life of the driver a whole lot easy. Enabling the driver to speak to his car infotainment system would allow him to control and navigate these functionalities with great ease. The car will be enabled with a virtual assistant. 4) Automatic rear view mirror adjustment scheme: Using the camera placed in front of the driver, the system shall detect the position of the driver’s head. This shall also be done using image processing techniques and we shall identify the coordinates of the driver head in 3D space. There will be a mapping between the head position and mirror adjustment scheme. The mirrors will adjust their position using servo motors and shall do so automatically by identifying the head position. 5) A revolutionary reverse-assistance algorithm for smart parking and general reversing: Probably the highlight of our model, this feature shall make driving the car in and out of a parking spot, or rather, even reversing a car in general, far easier and safer than what it already is. Like most other modern cars, our model shall also have a camera installed at the back and the corresponding image displayed at the infotainment screen for parking assistance. Upon activating the reverse gear, the screen shall trace the line of motion of the car corresponding to the current position of the steering wheel. Because of this feature, the driver gets an idea of whether or not he’ll hit an obstacle while reversing if the steering wheel is kept at that position. Taking this feature to another level, the rear camera, after capturing the live video feed from the back of the car, shall perform image processing and machine learning algorithms to find a safe, obstacle-free path for reversing and indicate the driver to move the steering wheel accordingly. So instead of relying solely on the drivers judgement, our system shall actually find the path to be taken while reversing, such that other cars and other obstacles will be avoided, and accordingly recommend the driver to steer the wheel in that direction. This feature shall be extremely useful for new drivers/ learners. During the initial phase, to prevent errors from creeping in, we will always have a manual override button. After a good amount of testing, further modifications and refinements can be made. Our systems adds new dimensions to both precautionary safety measures, as well as convenience. If implemented properly, we are confident that our project will reach new heights of HMI and driver assistance technology. It will give drivers several less reasons to worry about.

kaggle中的【State Farm Distracted Driver Detection】比赛的实现

Starter project for the Kaggle State Farm Distracted Driver Detection Competition

CNN model that detects distracted drivers on the road

Trained MobilenetV2 model to recognize the distracted behaviors exhibited by drivers while driving. Deployed the Deep Learning model on the flask to make real-time predictions.

kaggle走神司机预测

侦测走神司机

VGG & Resnet Neural Networks for Kaggle's State Farm Distracted Driver Detection contest (Tensorflow)

This repository implements distracted driver detection using a custom ResNet50 model with Leave-One-Group-Out cross-validation. It includes data preprocessing, training, and evaluation to classify driver behaviors across 10 categories.

State Farm Distracted Driver Detection Project using PyTorch

udacity机器学习工程师毕业项目

No description available

Distracted Driver detection with mobileVGG network on AUC dataset

This is a Python implementation of detecting mobile phones while driving using YOLO v5, performed using Kaggle's State Farm Distracted Driver Detection Dataset.

This Distracted Driver Detection Project is developed by a group of 5 students as part of "CS 539 Machine Learning" Course

Keep your eyes on the street while driving. This app alerts drowsy and distracted car drivers using face detection with camera. eyes2drive comes with an Apple Watch extension to show your trip events, to be able to start and stop your trip with via the Watch and most important get a Notification if a RED alert (long distraction) occured. eyes2drive is using state-of-the-art face detection algorithms using OpenCV algorithms to detect the face, eyes closing and head distraction.

Based on Kaggle Distracted Driver Detection, to do some try.

A CNN based Deep Learning Project to predict the behavior of the driver

An application to run on any device can be used as to track driver activity, whether a driver is Distracted or Drowsy and alert the Driver accordingly based on the alert meter. It tracks and detects various Driver activity such as Wearing Seatbelt, Using Phone, Where Driver is looking, eyes opened or closed and others.