Sara Mousavi

I am currently a Ph.D. student in the area of applied machine learning. During my Ph.D., I have developed an auto-curation framework for large image datasets to facilitate the utilization of such datasets for various research and industrial applications. My work has led to various novel unsupervised and semi-supervised computer vision techniques to support the auto curation. Throughout the design and development of this framework, I have used various tools, and languages such as Pytorch, Keras, Docker, Bash, and Python. Applying my presented auto-curation framework and the developed techniques on a domain-specific image dataset with more than 1M images resulted in a curated dataset and opened the path to research on several topics in the target domain that were not feasible otherwise.

• Designed, developed and implemented an auto-curation framework for large image data with evolving content
• Designed and implemented a semi-supervised semantic segmentation method for large image datasets with the limited labeled data available (Python, Pytorch, Docker, Opencv, Bash)
• Designed and implemented a semi-supervised method to enable mass image labeling while reducing human effort in the process (Python, Keras, Docker, Opencv, Bash)
• Designed and developed an unsupervised method for semantic clustering of large image collections with evolving content (Python, Keras, Docker, Opencv, Bash)
• Designed and implemented an MDS code to provide delayed parity generation with the goal of increasing write speed and avoiding network traffic while recovering lost data blocks in a data storage system (C).
• Implemented K-means clustering and PCA for analyzing a World Health Organization Dataset (Matlab).
• Implemented K-nearest neighbors and decision tree algorithms to classify benign and malignant breast tests in the Wisconsin breast cancer dataset (Matlab).
• Implemented a neural-network-based approach to classify spam emails from non-spam emails (Matlab and Python).
• Used random forest, bagging, boosting and classification algorithms in course projects (R).
• Implemented a Unix-like shell and file system using (C++).
• Utilized the Mapreduce parallel data processing system on the CloudLab platform (Hadoop, Spark and Python)
• Estimated the importance of Wikipedia articles using PageRank algorithm (Python).

• Sara Mousavi, Zhenning yang, Kelley Cross, Dawnie Steadman, Audris Mockus, “Similarity-based Label Reuse for Semi-Supervised Semantic Segmentation of Human Decomposition Images”, Submitted to IEEE Winter Conference on Applications of Computer Vision (WACV 2022).
• Sara Mousavi, Dylan lee, Tatianna Griffin, Dawnie Steadman, Audris Mockus, “SChISM: Semantic Clustering via Image Sequence Merging for Images of Human-Decomposition”, IEEE Winter Conference on Applications of Computer Vision (WACV 2021).
• Sara Mousavi, Dylan lee, Tatianna Griffin, Dawnie Steadman, Audris Mockus, “Collaborative learning of semi-supervised clustering and classification for labeling uncurated data”, IEEE International Conference on Image Processing (ICIP 2020).
• Tapajit Dey, Sara Mousavi, Eduardo Ponce, Tanner Fry. Bogdan Vasilescu, Anna Filippova, and Audris Mockus“Detecting and Characterizing Bots that Commit Code”, Accepted in Mining Software Repositories Conference, 2020.
• Sara Mousavi, Dylan lee, Tatianna Griffin, Dawnie Steadman, Audris Mockus, “An analytical workflow for clustering forensic images”, Extended Abstract, Thirty-fourth AAAI conference on artificial intelligence, NewYork, USA
• Sara Mousavi, Ramin Nabati, Megan Kleeschulte, Dawnie Steadman, Audris Mockus, “Machine-assisted annotation of forensic imagery”, IEEE International Conference on Image Processing (ICIP), Taipei, Taiwan
• Sara Mousavi, Chao Tian, Weizhe Zhang, “Delayed parity generation in MDS storage codes”, IEEE International Symposium on Information Theory (ISIT), Vail, Colorado
• Vahid Soleimani, Sara Mousavi, “MRI image segmentation using an improved Ant Colony optimization method”, 4th Iranian Conference on BioInformatics, Tehran

• Mentorship Chair, Systers: Women in EECS, University of Tennessee, Knoxville
• Travel Coordinator, Volhacks Hackathon, University of Tennessee, Knoxville

• EECS Departmental One-time Fellowship Award, University of Tennessee, Knoxville
• Grace Hopper Travel Award, Systers, Women in EECS at UTK
• EECS Departmental Excellence Fellowship, University of Tennessee, Knoxville
• Selected as top five BSc graduate, Computer Engineering Department, Razi University, Kermanshah
• Ranked top 5% in approximately 300,000 applicants of the National University Entrance Exam, Iran