Search Results - jayaraman+thiagarajan

9 Results Sort By:
Implementing Improved Diversity Using Adversarially Learned Transformations for Domain Generalization
Machine learning models and neural networks in particular can be used in the field of image processing. Machine learning models are trained to better improve their output quality. During training, domain generalization is the problem of making accurate predictions on previously unseen domains, especially when these domains are very different from...
Published: 2/12/2024   |   Inventor(s): Tejas Gokhale, Rushil Anirudh, Jayaraman Thiagarajan, Bhavya Kailkhura, Chitta Baral, Yezhou Yang
Keywords(s): Algorithm Development, Imaging, Machine Learning, Neural Computing, PS-Computing and Information Technology
Category(s): Computing & Information Technology, Physical Science
Multi-Layered Graph Embeddings for Machine Learning
The prevalence of relational data in several real-world applications, e.g., social network analysis, recommendation systems, and neurological modeling has led to crucial advances in machine learning techniques for graph-structured data. This encompasses a wide-range of formulations to mine and gather insights from complex network datasets — node...
Published: 4/5/2023   |   Inventor(s): Uday Shanthamallu, Andreas Spanias, Jayaraman Thiagarajan, Huan Song
Keywords(s): Machine Learning, Networks, IT, Software and Communication, PS-Computing and Information Technology
Category(s): Physical Science, Computing & Information Technology
Unsupervised Audio Source Separation Using Generative Priors
­Background Audio source separation, the process of recovering constituent source signals from a given audio mixture, is a key component in downstream applications such as audio enhancement and music information retrieval. Modern under-determined audio source separation systems rely on supervised training of carefully tailored neural network architectures...
Published: 2/23/2023   |   Inventor(s): Vivek Sivaraman Narayanaswamy, Jayaraman Thiagarajan, Rushil Anirudh, Andreas Spanias
Keywords(s): Networks, IT, Software and Communication
Category(s): Physical Science, Computing & Information Technology, Intelligence & Security
Audio Source Separation via Multi-Scale Learning with Dilated Dense U-Nets
Background Audio source separation refers to the process of extracting constituent sources from a given audio mixture. Despite being a critical component of audio enhancement and retrieval systems, the task of source separation is severely challenged by variabilities in acoustic conditions and the highly ill-posed nature of this inverse problem. A...
Published: 2/23/2023   |   Inventor(s): Vivek Sivaraman Narayanaswamy, Sameeksha Katoch, Jayaraman Thiagarajan, Andreas Spanias
Keywords(s): Networks, IT, Software and Communication
Category(s): Computing & Information Technology, Physical Science, Intelligence & Security
Deep Kernel Machine Optimization
Background The implementation of deep learning in a wide range of applications has resulted in an ever-increasing complexity of neural network architecture. These network architectures require reconfiguration to support specific applications, particularly in the field of computer vision where an image may contain a variety of objects needing to be...
Published: 2/23/2023   |   Inventor(s): Huan Song, Jayaraman Thiagarajan, Andreas Spanias
Keywords(s): Networks, IT, Software and Communication
Category(s): Computing & Information Technology
Triplet Network with Attention for Speaker Diarization
Background In automatic speech processing systems, speaker diarization is a crucial front-end component for separating speech segments by speaker without a priori knowledge about speaker identities. The first phase of many state-of-the-art diarization techniques involves the conversion of original speech data into representative i-vectors, which are...
Published: 2/23/2023   |   Inventor(s): Huan Song, Visar Berisha, Andreas Spanias, Megan Willi, Jayaraman Thiagarajan
Keywords(s): Networks, IT, Software and Communication
Category(s): Physical Science, Computing & Information Technology
Automated Tumor Segmentation
Counting the number of dots on an image has multiple applications. One such application is measuring the number of glomeruli (which appear as dots on images) in an intact kidney using non-destructive techniques to study renal and systemic failures. Some currently available techniques to count the glomeruli and measure their size require the destruction...
Published: 2/23/2023   |   Inventor(s): Jayaraman Thiagarajan, Karthikeyan Ramamurthy, Andreas Spanias, David Frakes
Keywords(s): Medical Devices and Imaging
Category(s): Computing & Information Technology
An algorithm to estimate glomerular number from kidney magnetic resonance images
Counting the number of dots on an image has multiple applications. One such application is measuring the number of glomeruli (which appear as dots on images) in an intact kidney using non-destructive techniques to study renal and systemic failures. Some currently available techniques to count the glomeruli and measure their size require the destruction...
Published: 2/23/2023   |   Inventor(s): Jayaraman Thiagarajan, Karthikeyan Ramamurthy, David Frakes, Andreas Spanias
Keywords(s): Medical Devices and Imaging, Networks, IT, Software and Communication
Category(s): Computing & Information Technology
Recovering Degraded Images using Ensemble Sparse Models
In many resource-constrained and noisy scenarios, the process of sensing as well as transmitting images introduces various forms of degradations. For example, in Magnetic Resonance Imaging (MRI), the lengthy sensing process may result in blurry images due to subtle movements of the patient. Bandwidth limitations and noisy channels may introduce degradation...
Published: 2/23/2023   |   Inventor(s): Karthikeyan Ramamurthy, Jayaraman Thiagarajan, Prasanna Sattigeri, Andreas Spanias
Keywords(s): Medical Devices and Imaging, Networks, IT, Software and Communication
Category(s): Computing & Information Technology