Jaypee Institute of Information Technology

About: Jaypee Institute of Information Technology is a education organization based out in Noida, Uttar Pradesh, India. It is known for research contribution in the topics: Computer science & Cluster analysis. The organization has 2136 authors who have published 3435 publications receiving 31458 citations. The organization is also known as: JIIT Noida.

Real Time License Plate Recognition from Video Streams using Deep Learning

Abstract: With ever increasing number of vehicles, vehicular management is one of the major challenges faced by urban areas. Automation in terms of detecting vehicle license plate using real time automatic license plate recognition (RT-ALPR) approach can have many use cases in automated defaulter detection, car parking and toll management. It is a computationally complex task that has been addressed in this work using a deep learning approach. As compared to previous approaches, license plates have been recognized from full camera stills as well as parking videos with noise. On a dataset of 4800 car images, the accuracy obtained is 91% on number plate extraction from images, 93% on character recognition. Proposed ALPR system has also been applied to vehicle videos shot at parking exits. Overall 85% accuracy was obtained in real-time license number recognition from these videos.

Evolving large scale healthcare applications using open standards

Abstract: Electronic Health Records (EHRs) are becoming more prevalent in health care. Worldwide exchange of healthcare data demands adherence to semantic interoperable standards to overcome the language and platform barriers. Various healthcare organizations in developing countries such as, India adopt their own independent information systems without adhering to standard guidelines. Thus, this tends to sacrifice interoperability. This affects permanent persistence of longitudinal health records for future reference and research purpose. Current research implements a standard based clinical application to be used for healthcare domain in India. The study has been done for enhancing the data quality through standardization. It aims at providing a generic permanent persistence to track life-long interoperable health records of patients. This is the first effort for exploring its adoption for various regional languages in India. The user interfaces have been generated for various Indian languages for testing on a sample set of archetypes. The clinical application deployed in ‘Hindi’ language can be easily deployed for other people in ‘Tamil’ language, while maintaining semantic interoperability. The persistence will also be maintained, with the same meaning (of data) for both the regions. Implementing these standard based healthcare applications helps in reducing the costs while enhancing patient care. Thus, this study aims to build a standard based, and platform independent healthcare application to provide support for interoperability, usability and generic persistence.

PAPR reduction of OFDM systems using PTS with genetic algorithm at low computational complexity

Abstract: Orthogonal frequency division multiplexing (OFDM) is one of the preferred choices for high speed transmission of data in wireless domain pertaining to its immunity towards inter-symbol interference. However high value of peak to average power ratio (PAPR) is a major problem associated with OFDM. High PAPR increases circuit complexity and reduce RF amplifier efficiency. Partial Transmit Sequence (PTS) is one of the widely used techniques for PAPR reduction, however in PTS scheme the computation of optimal phase factors necessitates exhaustive searching among all allowable phase factors. This leads to exponential increase in search and computational complexity with the increase in number of subblocks. In this paper a PTS scheme using Genetic algorithm (GA) for optimized search with low computational complexity scheme is proposed. The proposed scheme provides efficient search and PAPR performance at low computational complexity for a given value of PAPR threshold.