Bharati Vidyapeeth's College of Engineering

About: Bharati Vidyapeeth's College of Engineering is a based out in . It is known for research contribution in the topics: Deep learning & Computer science. The organization has 709 authors who have published 622 publications receiving 3550 citations.

Visual Assistant for Blind People using Raspberry Pi

Abstract: Blind people face the problem in daily life. They can't even walk without any aid. Many times they rely on others for help. Several technologies for the assistance of visually impaired people have been developed. Among the various technologies being utilized to assist the blind, Computer Vision-based solutions are emerging as one of the most promising options due to their affordability and accessibility. This paper proposes a system for visually impaired people. The proposed system aims to create a wearable visual aid for visually impaired people in which speech commands are accepted by the user. Its functionality addresses the identification of objects and signboards. This will help the visually impaired person to manage day-to-day activities and navigate through his/her surroundings. Raspberry Pi is used to implement artificial vision using python language on the Open CV platform.

BlockFITS: A Federated Data Augmentation Modelling for Blockchain-Based IoVT Systems

Abstract: In Intelligent Transport Systems (ITS), the collection of diverse data is a major practical roadblock; not only can their data be personally identifiable, i.e. private, but also the lack of incentive for entities to participate in any kind of collaborative training is also severely limited due to the added computational expense of training collaborative models locally. In this paper, we propose BlockFITS: A Vehicle-to-BlockChain-to-Vehicle (V2B2V) federated learning enabled model training paradigm for ITS entities. In addition to which we propose a data augmentation scheme that operates with cooperative training to generate an incentive for entity participation. The immutability and decentralised features of the Blockchain system leverage the federated-like averaging of synthetically generated data samples that generate incentives for the participation of entities in such a training setup. BlockFITS can be practically deployed in future ITS systems to improve the autonomous driving system, pedestrian safety, and vehicular object detection or more due to its model-constraint-free characteristics which provide access to a synthetic and global data whilst maintaining data privacy.

Implementation of PFSCL demultiplexer

Abstract: A novel implementation of a demultiplexer in positive feedback source-coupled logic (PFSCL) style is presented in this paper. Two methods to realize PFSCL demultiplexer are described. The first one represents uses the conventional NOR based approach while the fundamental cell based approach is applied in the second one. The functionality of the proposed demultiplexer based on the two approaches is validated through using TSMC 180 nm MOS parameters. A comparison in the performance of the PFSCL demultiplexers is also performed in terms of various parameters. It is found that the fundamental cell based demultiplexer is better than the NOR based PFSCL demultiplexer.

Adiabatic logic based circuit optimization for ultra low power applications

Abstract: In this paper authors have presented a design and experimentation evaluation of circuit optimization inspired by adiabatic logic processing. The reported work inherent the power benefits as...

Virtual Autonomous Vehicle Using Deep Reinforcement Learning

Abstract: In this paper, we propose an in-depth reinforcement learning approach to evaluate the performance of the virtually created autonomous vehicle driving scenario. Markov Decision Process is used to map the state of the vehicle to an action. Discount and Reward functions are also incorporated in the decision policy. To deal with high-dimensional inputs which lead to the standard instabilities of reinforcement learning, we have used experience replay. To further reduce the correlation, we use an iterative update that updates the Q-values periodically. Adam Optimizer, based on stochastic objective functions, is used as an optimizer in a neural network along with Rectified Linear Unit activation function, which helps in further optimizing the process. The autonomous vehicle doesn't need any labelled training data to learn human driving behaviour. Inspired by the real-world scenarios, an action-based reward function is used to train the vehicle. It has been shown in our approach that after some iterations, the virtually created vehicle generates collision-free movement and performs the same driving behaviour as of humans.