Objective | Problem Statement: To detect driver distraction using the in-car images using machine learning approaches.
Motivation: Road safety is one of the major concerns in India. According to Global Status report [1] on Road safety 2018, more than 1.35 million people were killed across the world with 11% causalities reported from India alone which is highest among 199 countries. This work proposes an approach to detect visual, manual and cognitive real time driver detection using the image processing and machine learning approach.
Dataset used for Evaluation: The American University in Cairo distracted driver (AUC) dataset [2]. This dataset contains around 12978 training images and around 4330 testing/validation images showing the driver’s behaviours. Some sample images along with their class are shown in fig 1. The other similar dataset which may be used for evaluation is Statefarm dataset available on Kaggle[3]
References:
1. WHO Global status report on road safety 2018 https://www.who.int/publications/i/item/9789241565684
2. Y. Abouelnaga, H. Eraqi, and M. Moustafa. ”Real-time Distracted Driver Posture Classification”. Neural Information Processing Systems (NIPS 2018), Workshop on Machine Learning for Intelligent Transportation Systems, Dec. 2018. https://arxiv.org/abs/1706.09498.
3. Statefarm driver distraction dataset, https://www.kaggle.com/c/state-farm-distracted-driver-detection. |