Projects

Keywords: COVID-19, coronavirus, SARS-CoV-2, disease transmission, weather, Google Trends, air pollutants, economic indicators, Infodemiology

Research Question(s): How can we use a large-scale, multi-variable analysis to identify the critical factors for COVID-19 transmission?

Description: Since the emergence of Covid-19 in early 2020, we have all been battling to stay safe, infection free, and most importantly to avoid spreading the disease to friends, loved ones, and strangers alike.  Understanding the important factors that contribute to Covid-19 transmission is an important task for the general public, and public health officials to help stop the continuation of the global pandemic, deaths related to Covid-19, and the economic fallout which will ensue if we cannot return to regular life.

Keywords: Public health, overweight, food, language, Twitter, deep learning

Research Question(s): How can we utilize deep learning natural language processing models on social media to predict and find patterns between online data and overweight users?

Description: Obesity is one of the highest health concerns in the United States and is of upmost importance to public health officials to reign in the possible causes, correlations, and determining factors.  In this study, using self-disclosure theory, which describes people having the tendency to disclose otherwise personal information more freely online than they would in person, we attempt to predict if a given user is overweight or not. 

Keywords: COVID-19, COVID-19 vaccine, sentiment evolution, topic modeling, social media, text mining

Research Question(s): What are the general sentiments on COVID-19 vaccines? What are the topics that shape the sentiments? How do concerns (ie, topics with negative sentiments) evolve over time?

Description: Sentiment and topic modeling is an important task machines have learned to conduct very well and is a useful tool for public officials in charge of policy decisions.  The information returned by these models can offer complex analysis and powerful data to be used as a tool to improve the effectiveness of the decision-making process for those tasked with such responsibility.  This study explores the sentiment and topics surrounding Covid vaccines and gives some important insight on how one may choose to use such information for policy decisions. 

Keywords: Thorax Disease, X-Ray, Knowledge-Guided deep Zoom Neural Network, Weakly-supervised learning, Chest X-Ray image,  Computer-Aided Design 

Research Question: How can we use CNN for Thorax Disease detection, and can we get results better than the regular methods?

Description: Diagnosis of thorax disease automatically and accurately is important for clinical assist analysis. But the major challenge is noise and similarity between visual features. For this we used KGZNet, which is a knowledge-guided deep zoom neural network. KGZNet is a data driven model. So this model learns from prior knowledge and finer region-based feature represen tation. Our results show that lung and lesion regions can boost the performance of thoracic disease classification and reinforce each other. 

Keywords: Pavement distress detection, Convolutional Neural Networks, Expectation-Maximization algorithm, image classification, object localization

 Research Question:  How can we detect Pavement Distress Iterative Optimized Patch Label Inference Network? 

Description: Potholes and other pavement disease is rampant throughout the vast networks of roadways that now connect our cities.  It quickly becomes a daunting task to identify areas of distress in the countless miles of roadways the department of transportation is given charge of.  In this paper, we explore the implementation of CNNs to not only identify, but classify, pavement disease.  This is an extremely useful tool for those professionals given the task of identifying hot-spot regions which require attention from maintenance crews. 

Keywords: Histograms, Shape primitives histograms, 3D face recognition, 2D face recognition, Multi-Scale shape primitives histogram

Research Question: How can we perform face recognition using shape primitives of face?

Description:  It’s hard to ignore the fact that facial recognition is one of the key areas in machine learning helping with improvements in security, efficiency, and entertainment.  As such, recognizing the deficits of modern state-of-the-art learning algorithms is an important task as we move towards the future.  In this paper, we utilize low-level features which are gathered by partitioning facial images into numerous smaller shape fragments, which are then reduced themselves into several atomic shape patterns called “shape primitives”.  By improving state-of-the-art models by introducing this technology, we help usher in the modern age of facial recognition.

Keywords: Cancer, out of hospital, hierarchical linear linear regression. 

Research Question: Can the duration of time between two consecutive admissions play an important role in the treatment of cancer?

Description: Cancer is a lethal disease. Duration of time between two consecutive admissions or out of hospital plays an important role in healthcare service quality. We used data of 22,231 admissions. Methodology used was hierarchical linear regression on factors like age, marital status, number of admissions and whether the treating hospital is in the same province as the patient. 

Keywords: Corner Detection, Centroid, point-to-centroid, edge detection.

Research Question: How can we detect corners using the centroid location?

Description: Corners, highly important local features of images and corner finding, play a crucial role in computer vision and image processing, such as object tracking and vehicle detection. Proposing effective and efficient corner detectors is the aim of corner detection. In this study, the authors first present a new measure of corner sharpness termed as the point-to-centroid distance (PCD) and then examine its behaviours, which display beneficial characteristics that help distinguish corners from non-corners.   

And Many More Projects yet to come.......