PROJECT TITLE :

Deep Learning-Based Vehicle Behavior Prediction for Autonomous Driving Applications A Review

ABSTRACT:

The Behaviour Prediction Function of an Autonomous Vehicle is Responsible for Making Predictions Regarding the Future States of Other Vehicles in the Area Based on Current and Previous Observations of the Surrounding Environment. This helps them become more aware of the potential dangers that are nearby. However, conventional solutions for behavior prediction can be used in straightforward driving situations that call for limited time horizons for their predictions. In recent years, deep learning-based strategies have gained popularity as a result of their promising performance in more complex environments when compared to conventional strategies. This is due to the fact that deep learning strategies were originally developed by Google. In this article, we provide a comprehensive review of the state-of-the-art of deep learning-based approaches for predicting how a vehicle will behave. Our motivation for doing so comes from the growing popularity of this technique. First, we provide an overview of the more general problem of predicting vehicle behavior and discuss the difficulties associated with it. Next, we classify and review the most recent deep learning-based solutions based on three criteria: the representation of the input data, the type of output data, and the method of prediction. Additionally, the performance of several well-known solutions is discussed in this article, along with the identification of research gaps in the existing literature and an outline of potential new directions for research.