This is the code of paper entitled "AFM3D: An Asynchronous Federated Meta-learning Framework for Driver Distraction Detection".

Implemantion of a lightweight neural network architecture for the detection of distracted driving among drivers.

An academic project that implements a Convolutional Neural Network (CNN) entirely from scratch using NumPy — no PyTorch or TensorFlow. Designed to detect driver distractions from images and videos, it includes a Streamlit frontend for real-time testing. Built to understand deep learning fundamentals through hands-on coding.

Certainly! The Head-Eye Tracker model consists of several components and functionalities to enable head pose estimation, eye tracking, glasses detection, gaze estimation, and intersection with a 3D representation of the car's interior.

A simplified and improved academic project implementing a CNN from scratch using NumPy to detect driver distractions. D3-CNN-2 achieves ~86% test accuracy in just 11 epochs, outperforming the original model with fewer layers. Includes a Streamlit-based frontend for real-time image/video testing — all without using PyTorch or TensorFlow.

Real-time driver distraction detection using time-distributed convolutional LSTM network for mobile platforms

Driver Distraction Detection with CNN and Transfer Learning (VGG19, EfficientNet)

Infothon 3.0

Project for "Computer Vision and Cognitive Systems" course @ Unimore

[ICASSP 2025]RAPID: Recognition of Any-Possible DrIver Distraction via Multi-view Pose Generation Models

This work was supported in part by the MOTIE (Ministry of Trade, Industry & Energy), Republic of Korea, under the Technology Innovation Program, and in part by the MSIT (Ministry of Science and ICT), Republic of Korea, under the Grand ICT Research Center Support Program.For full details, refer the published journal article using the link below.