Researcher in Computer Vision, Conservation, and Sustainability
Hi! I'm a third year PhD student in Computing and Mathematical Sciences at Caltech, advised by Pietro Perona. The natural world has always resonated with me with its intricate beauty, tranquility, and the depth of scientific exploration it offers.
After six years as a software engineer, I plunged into the research world - an exciting and transformative transition! I am now an active part of my research group. I focus on computer vision and machine learning approaches that can have broad scientific, societal, and environmental impacts.
My research involves developing robust computer vision methods applicable to biodiversity and environmental monitoring across various data modalities. I aim to ensure that my research is adaptable to new environments with limited resources, while dealing with challenges such as domain shifts, noisy data, and long-tailed distributions, inherent in all fine-grained problems. My ultimate goal is to produce research that can contribute to effective policy-making for ecological protection.
In line with my personal beliefs, I firmly uphold that the fields of Science, Technology, Engineering, and Mathematics (STEM) should be universally accessible, irrespective of one's background, gender, race, nationality, sex, age, or religion. My experience as a History major and a nontraditional graduate student have taught me lifelong skills and shaped my perspectives. I value the synergy of interdisciplinary perspectives and understand the significant impact that precise computational techniques can have on the world around us.
Check my linkedin for a more detailed list
2017-2021, Amazon
Software Development Engineer
With a team of roughly 4 other developers, I developed a cross-platform C++ music player aimed to improve customers’ listening experiences.
Playback library was developed to integrate with iOS, Android, Desktop (Mac/PC), Amazon Fire, and more.
2015-2017, NASA JPL
Engineering Applications Software Engineer
Worked with the Cyber Defense Engineering and Research Group on a tool used
to analyze the cybersecurity landscape of mission software systems (C++, Python)
Generated pairwise features from website forum metadata on the dark web to boost precision
of user-pair identification. This work was leveraged to identify human traffickers (Python)
This includes some of my recent publications. Check Google Scholar for a more complete list
Stathatos, Suzanne, Asitang Mishra, Chris A. Mattmann. Cyber Persona Identification via Indirect Feature Analysis, WSDM 2018
Pecharich, Jeremy, Suzanne Stathatos, Brian Wright, Arun Viswanathan, Kymie Tan. Mission-Centric Cyber Security Assessment of Critical Systems, AIAA 2016
Won, A.S., Bailenson, J.N., Stathatos, S.C. et al. Automatically Detected Nonverbal Behavior Predicts Creativity in Collaborating Dyads. J Nonverbal Behav 38, 389–408 (2014)
Suzanne Stathatos is a third year PhD student studying Computing and Mathematical Sciences. With Pietro Perona's lab, she leverages and develops computer vision methods to enable biodiversity monitoring across data modalities and solve real-world sustainability challenges. These, for example, include tackling challenges in geospatial and temporal domain shifts, learning from noisy data and long-tailed distributions, and fine-grained categorization. Prior to Caltech, she held software engineering positions at Amazon Music and JPL, which Caltech manages for NASA. Stathatos earned her bachelors in history and masters in computer science at Stanford University.
I am a member of the Pacific Council. There, I get to understand and connect with policy experts to help turn my research into reality, as I hope to ultimately implement sweeping advances in wildlife conservation.
As you can see from my photo up top, I enjoy any excursion outside - hiking, camping, mountaineering (favorite mountain is Mount Shasta).
In college, I was a history major focusing on wartime technology advancements. For my senior capstone project, I wrote about how the Claude Chappe semaphore, a precursor to the telegraph, affected and was influenced by the political and socioeconomic landscape in which it arose.
