Browsing by Author "Thomas, E."

Filter results by typing the first few letters
Now showing 1 - 6 of 6
  • No Thumbnail Available
    Item
    An Improved and Reliable Sequential Decoding of Convolution Codes
    (Institute of Electrical and Electronics Engineers Inc., 2020) Chandavarkar, B.R.; Byju, A.; Thomas, E.
    Error control (detection and correction) of data plays a pivotal role in networking to facilitate the reliable transmission of messages from source to destination. Convolution code is one of the popularly known error control mechanism which is considered superior to several legacy error control algorithms like the Hamming code. The Fano algorithm is a sequential decoding algorithm used to decode long constraint length convolution codes. However, this algorithm has failed to offer 100% error detection and correction capabilities. These lead to the inferior performance of this algorithm to correct errors at the destination. This paper proposes an enhanced Fano (e-Fano) algorithm that offers 100% error detection and correction up to finite bits of error. Through the MATLAB simulations, e-Fano and the conventional Fano algorithm are compared for the % of error detection and correction in the received data. © 2020 IEEE.
  • No Thumbnail Available
    Item
    Distributed Cloud Deep Learning Architecture for Complex Image Analysis and Run-time Prediction Tool
    (Springer Science and Business Media Deutschland GmbH, 2021) Kumar, S.; Thomas, E.; Horo, A.; Annappa, B.
    Hyperspectral imaging is a rare research tool and has been transformed into a commodity product found in a wide field. Currently, standard data processing methods that specialize in special hyperspectral accumulation structures are required. Also, with the advent of data collection and development in the field of sensory devices, it has rendered previous processing tools in vain. To manage this huge increase in the amount of data, a consistent cloud distribution method is required. Hyperspectral images (HSIs) have several spectral band channels that make the study very difficult. In this paper, an in-depth reading method of the novel with a modified autoencoder is proposed as a cloud-based use of HSI analysis, which provides a measure of lesser error rates and high accuracy of classification models. In line with this, a list of four tools has been proposed to calculate the actual number of workers, cores, and iterations required to achieve the desired accuracy for a specified amount of run-time. This will help cloud managers get a basic idea of computational needs and help them allocate resources more efficiently. The entire architecture was simulated on Spark servers and was verified experimentally by checking that the proposed architecture performs the function of efficient management and analysis of large HSI. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
  • No Thumbnail Available
    Item
    Identifying Parking Lots from Satellite Images using Transfer Learning
    (Institute of Electrical and Electronics Engineers Inc., 2019) Kumar, S.; Thomas, E.; Horo, A.
    With the advent of digital image processing techniques and convolutional neural networks, the world has derived numerous benefits such as computerized photography, biological Image Processing, finger print and iris recognition, to name a few. Computer vision coupled with convolutional neural networks has attributed machines with a virtual intellectual ability to recognize and distinguish images based on several characteristics that may be impossible for the human eye to perceive. We have exploited this advancement in technology to particular use case of detecting number of empty and occupied parking slots from satellite images of parking lots. We have proposed a befitting sequence of classical image processing techniques and algorithms to perform pre-processing of satellite images of parking spaces. Moreover, we have proposed a Convolutional Neural Network model that takes as input these preprocessed images and identifies the empty and occupied parking slots with an accuracy of 97.73%. The potential benefits of using Neural Networks to realize the objective can be extended to open parking spaces of different configurations. This is due to the fact that establishing sensors over a large number of parking slots over a given open parking space can be a cumbersome and exorbitant task. The proposed model comprises of few convolutional layers and uses Rectified Linear Classification activation function. © 2019 IEEE.
  • No Thumbnail Available
    Item
    Logistic regression based DFS for Trip Advising Software (ASCEND)
    (2019) Thomas, E.; Byju, A.; Chandrasekaran, K.; Usha, D.
    Graphs have played a pivotal role in the field of computer science and has been an efficient method for representing and modeling abstractions in various fields. They can be used to represent several real life models. Several domains in today's world use the concept of graphs extensively such as GPS Navigation systems, Computer networks, WebCrawler, Social Networking websites, peer to peer networking, medical and biological field, neural networks etc. Taking into account the numerous applications of the concept of graphs in today's world, graph searching becomes inevitably significant. In this scenario it is important to note that several graph searching algorithms that were proposed to give exhaustive searches doesn't provide the most satisfying outcome in terms of asymptotic time complexity. Through this paper we intend to highlight the significance of machine learning as a useful tool that can be incorporated in various graph searching algorithms that can reduce its complexity. We classify the existing graph searching techniques as subsets or modifications of two major conventional graph searching algorithms namely BFS(Breadth First Search) and DFS(Depth First Search) and suggest the application of logistic regression to improve their performance. It is confounding that only few research papers explore the application of machine learning to the aforementioned graph searching algorithms. Hence, it is evident that there exists scope for future research on this topic and we aim to suggest directions for the same. � 2019 IEEE.
  • No Thumbnail Available
    Item
    Logistic regression based DFS for Trip Advising Software (ASCEND)
    (Institute of Electrical and Electronics Engineers Inc., 2019) Thomas, E.; Byju, A.; Chandrasekaran, K.; Usha, D.
    Graphs have played a pivotal role in the field of computer science and has been an efficient method for representing and modeling abstractions in various fields. They can be used to represent several real life models. Several domains in today's world use the concept of graphs extensively such as GPS Navigation systems, Computer networks, WebCrawler, Social Networking websites, peer to peer networking, medical and biological field, neural networks etc. Taking into account the numerous applications of the concept of graphs in today's world, graph searching becomes inevitably significant. In this scenario it is important to note that several graph searching algorithms that were proposed to give exhaustive searches doesn't provide the most satisfying outcome in terms of asymptotic time complexity. Through this paper we intend to highlight the significance of machine learning as a useful tool that can be incorporated in various graph searching algorithms that can reduce its complexity. We classify the existing graph searching techniques as subsets or modifications of two major conventional graph searching algorithms namely BFS(Breadth First Search) and DFS(Depth First Search) and suggest the application of logistic regression to improve their performance. It is confounding that only few research papers explore the application of machine learning to the aforementioned graph searching algorithms. Hence, it is evident that there exists scope for future research on this topic and we aim to suggest directions for the same. © 2019 IEEE.
  • No Thumbnail Available
    Item
    Multi-Res-Attention UNet: A CNN Model for the Segmentation of Focal Cortical Dysplasia Lesions from Magnetic Resonance Images
    (Institute of Electrical and Electronics Engineers Inc., 2021) Thomas, E.; Pawan, S.J.; Kumar, S.; Horo, A.; Niyas, S.; Vinayagamani, S.; Kesavadas, C.; Rajan, J.
    In this work, we have focused on the segmentation of Focal Cortical Dysplasia (FCD) regions from MRI images. FCD is a congenital malformation of brain development that is considered as the most common causative of intractable epilepsy in adults and children. To our knowledge, the latest work concerning the automatic segmentation of FCD was proposed using a fully convolutional neural network (FCN) model based on UNet. While there is no doubt that the model outperformed conventional image processing techniques by a considerable margin, it suffers from several pitfalls. First, it does not account for the large semantic gap of feature maps passed from the encoder to the decoder layer through the long skip connections. Second, it fails to leverage the salient features that represent complex FCD lesions and suppress most of the irrelevant features in the input sample. We propose Multi-Res-Attention UNet; a novel hybrid skip connection-based FCN architecture that addresses these drawbacks. Moreover, we have trained it from scratch for the detection of FCD from 3 T MRI 3D FLAIR images and conducted 5-fold cross-validation to evaluate the model. FCD detection rate (Recall) of 92% was achieved for patient wise analysis. © 2013 IEEE.