Fear Learning Hardware Models for Intelligent Robots

Assignment Task

Summary of research proposal for Vice- chancellor scholarship

Terms such as artificial intelligence, data science, deep learning, and neural networks are very common nowadays. In recent years, artificial intelligence has brought notable changes to almost every field. It will be a field of education or a field of medical science, or it will be a field of the manufacturing industry or a research field. Recent advances in artificial intelligence seek to incorporate emotions into machine learning models. Electronic systems' flexible and adaptive artificial intelligence has been greatly influenced by fear learning. The research goal is to create robot movement that responds to the frightening incident and to investigate the performance of neural network-based autonomic controllers and their adaptation to cope with scary situations.

Introduction

The aim of artificial intelligence is to mimic human intelligence and solve real-life problems. Artificial intelligence works similarly to the human brain by replicating the workings of neurons present in the human brain. The human brain learns things by observing them and analyzing them on the same ground that machines will learn decision-making. and this process is called training in terms of artificial intelligence. If all the terms related to artificial intelligence are arranged in a sequence, then the neural network will come first. They are the algorithms used for deep learning. [1] Deep learning is a method to train machines for decision-making, and that process is called machine learning. Machine learning is a subset of artificial intelligence; thus, all of the terms are interconnected. To implement machine learning algorithms Good computational power is required. along with the large data sets. A machine learning model will be trained with the dataset, and then it will produce the output from the given input. Robotics is one of the major applications of artificial intelligence. Robots are capable of understanding human language and performing tasks on their own with the help of machine learning and artificial intelligence. Robots can have complex structures like human beings. They are equipped with sensors such as cameras and motion sensors so that they can analyze all the inputs from their surroundings and perform the appropriate action as per the training given to them. On the other hand, robots can be simple objects that will just perform repetitive tasks. Be it a complex robot or a simple robot, they both need training tomake decisions. Artificial intelligence has made robots as intelligent as human beings, but they share only one aspect of human behavior, and that is emotions. If machine learning models are capable of learning or understanding the emotions of humans, then human-machine interaction will happen in a much more natural way, similar to that of human-human interactions. [2] Machine learning algorithms are extremely complex and require a lot of computing power. Multicore CPUs are a good platform to run those models and train them. But CPUs perform sequential operations, and to achieve good performance, those sequential operations can be divided across multiple cores; going ahead, GPUs (graphics processing units) are getting used for machine learning. GPUs perform parallel processing; they divide large tasks into smaller subtasks, and those subtasks will be performed by a specific number of cores available on that GPU. FPGAs have also shown good development, and they can be used for deep learning algorithms.

Literature Review

According to the research in [6], during the pandemic, online learning was the primary mode of instruction for all types of students, whether they were college students or high school students. During online learning, students will be present in front of the camera, but different studies show that this learning is not as effective as classroom learning. To make online learning very effective, it is very important that teachers know the emotional state of the student. But in the case of online learning, it is really impossible to look at every student continuously and get acknowledgement that they were able to understand or not. With the help of images captured of students' faces, one can get an idea of the emotional state of that student, and that will make the online learning process very easy for teachers. In this paper, student expressions such as excitement, confusion, and hardness are analyzed with a neural network.

As per [7], an end-to-end convolutional neural network was designed with the help of the Intel FPGA SDK for OpenCL targeting Intel FPGAs. This paper also focuses on optimizing high-level synthesis design while using it for FPGAs. It was also demonstrated that FPGA-designed convolutional neural networks are more power efficient than CPU and GPU-designed convolutional neural networks.

Aim and Objective

Research and development in the field of artificial intelligence has progressed so well that decisions made by machine learning models are as accurate as decisions or predictions made by human beings. Self-driving cars are one of the best examples of technology replacing the presence of human beings. The only thing missing is an emotional intelligence model for machines. The aim of the proposed research project is to design a machine learning model that can predict emotional behavior, and fear is the emotion under consideration. The aim is to integrate such emotional intelligence into robots so that they will behave the same as humans in unexpected situations. Those unexpected events are the source of fear. For running design models, there are multiple processing units available, such as a GPU, CPU, or any other programmable chip.

This SOP has been solved by our PhD Experts at My Uni Paper.