Conference Papers
Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506
Browse
5 results
Search Results
Item Detecting Semantic Similarity of Documents Using Natural Language Processing(Elsevier B.V., 2021) Agarwala, S.; Anagawadi, A.; Reddy Guddeti, R.M.The similarity of documents in natural languages can be judged based on how similar the embeddings corresponding to their textual content are. Embeddings capture the lexical and semantic information of texts, and they can be obtained through bag-of-words approaches using the embeddings of constituent words or through pre-trained encoders. This paper examines various existing approaches to obtain embeddings from texts, which is then used to detect similarity between them. A novel model which builds upon the Universal Sentence Encoder is also developed to do the same. The explored models are tested on the SICK-dataset, and the correlation between the ground truth values given in the dataset and the predicted similarity is computed using the Pearson, Spearman and Kendall's Tau correlation metrics. Experimental results demonstrate that the novel model outperforms the existing approaches. Finally, an application is developed using the novel model to detect semantic similarity between a set of documents. © 2021 Elsevier B.V.. All rights reserved.Item Molecular-InChI: Automated Recognition of Optical Chemical Structure(Institute of Electrical and Electronics Engineers Inc., 2022) Kumar, N.; Rashmi, M.; Ramu, S.; Reddy Guddeti, R.M.With the advent of a new era dominated by digital media and publications in recent years, the importance of striking a balance between traditional and new modes of operation has become increasingly apparent. It has been standard practice in the field of chemistry for decades to express chemical compounds using their structural forms, referred to as the Skeletal formula. In this research, we tried to interpret these old chemical structure images, extracted from old literature, to transform pictures back to the underlying chemical structure labeled as InChI text using a huge set of synthetic image data produced by Bristol-Myers Squibb. In this paper, we propose an improved synthetic data and an Encoder-Decoder-based deep learning-based model to automatically represent these molecular images into their underlying InChI representation. © 2022 IEEE.Item Fall Detection and Elderly Monitoring System Using the CNN(Springer Science and Business Media Deutschland GmbH, 2023) Reddy Anakala, V.M.; Rashmi, M.; Natesha, B.V.; Reddy Guddeti, R.M.Fall detection has become a critical concern in the medical and healthcare fields due to the growing population of the elderly people. The research on fall and movement detection using wearable devices has made strides. Accurately recognizing the fall behavior in surveillance video and providing the early feedback can significantly minimize the fall-related injury and death of elderly people. However, the fall event is highly dynamic, impairing categorization accuracy. The current study sought to construct a fall detection architecture based on deep learning to predict falls and the Activities of Daily Living (ADLs). This paper proposes an efficient method for representing extracted features as RGB images and a CNN model for learning the features needed for accurate fall detection. Additionally, the proposed CNN model is used to test for and locate the target in video using threshold-based categorization. The suggested CNN model was evaluated on the SisFall dataset and was found to be capable of detecting falls prior to impact with a sensitivity of 100%, a specificity of 96.48%, and a response time of 223ms. The experimental findings attained an overall accuracy of 97.43%. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.Item A Key-frame Extraction for Object Detection and Human Action Recognition in Soccer Game Videos(Institute of Electrical and Electronics Engineers Inc., 2023) Chopra, H.; Mundody, S.; Reddy Guddeti, R.M.In professional team games, sports analysts frequently analyze to learn tactical and strategic insights into the actions of players in these team games. The foundation of current analytic procedures is the examination of team footage. We provide a visual analytical framework that seamlessly combines abstract visualizations with team sports video recordings. It offers an exciting opportunity because several complicated, real-time occurrences are examined towards making strategic decisions. Visual object detection is a well-known and active research area. Any object, its speed, and its appearance have their level of detection difficulty in the face of numerous obstacles. Human Action Recognition (HAR) is required to carry out advanced operations in team games as there is an increase in demand for video analysis of sporting events. To strategically improve the team's performance, the team coach may, for instance, use an automatic monitoring system to monitor the player's movement and locations throughout a soccer match as well as the location of the football. This paper proposes a YOLOv7 model that uses the key-frame selection technique to analyze players' actions during a soccer game. In addition to detecting the football, player, and referee, the deep learning model can recognize six of the human actions in the soccer game. The experimental results show that using the key-frame selection technique for human action recognition, the total execution time can be reduced by approximately 68% to 70%. © 2023 IEEE.Item An Effective Early Detection and Prediction System for Gas Leakage in Smart Environments(Institute of Electrical and Electronics Engineers Inc., 2023) Ekka, N.; Mundody, S.; Reddy Guddeti, R.M.Gas leakages can be catastrophic, resulting in human injuries and financial losses. If the gas leaks can be detected and predicted before time, it can significantly help prevent any hazards. This paper proposes to develop a gas leakage detection system using reliable techniques to avoid such situations. The key objective of this paper is to develop a detection and prediction method to identify gas leak situations and predict the amount of gas released and its concentration by the time of release. A sensor-based approach and the Internet of Things (IoT) are employed to find gas leaks in enclosed spaces. For tasks involving detection and prediction, deep learning methods like Long Short-Term Memory (LSTM) networks are used. For evaluation purposes, this paper also compares the suggested strategy with other state-of-art techniques. Additionally, a monitoring and alert system is developed to notify users about gas leakage and hazards. © 2023 IEEE.