Projects 👩🏻‍💻

Position Estimation and Subpixel Accuracy in Tracking Particles from Images: I investigated the phenomenon of pixel locking, an error that limits the tracking accuracy of experimental outputs to a much lower resolution than is desired. Existing algorithms to track particles infer subpixels with a systematic error known as pixel locking, a phenomenon that biases the positions of particles towards certain locations, thereby limiting the tracking accuracy of experimental outputs to a much lower resolution than is desired. The objective of the project was to train a machine learning model to track the centers of particles with better precision to reduce the pixel locking error.

All isn’t unbiased in Climate and Violence Reportage: We studied media bias in the climate crisis. The study backs up the media-created barriers that inhibit the public from having a more comprehensive awareness and grasp of socio-political events. Our conclusion was that misinformation is one of the factors that shape the trajectory of discourse on climate change. Echo chambers—combined with confirmation bias and polarization—exacerbate the exposure to climate misinformation. The paper can be viewed here. (Collaborator: S. De)

Homestead Green: Designing the Buildings of Tomorrow Designed an energy-efficient skyscraper for both residential and commercial use with a low carbon footprint on the environment, featuring all modern conveniences that enhance the well-being of occupants while merging nature and technology to promote a green lifestyle. The skyscraper integrates various sustainable technologies and green walls which will contribute to the fauna adaptation in the urban environment. Press mentions: Fast Company, New York Academy of Sciences. (Collaborators: D. Filip, B. Singh, I. Tarabasanu, S. Dangi, I. Shafkat)

Podible Created a solution to effectively combat malnutrition among expectant mothers and children under the age of five in India. Podible is a cheap but impactful dissolvable nutrient pod that acts as an effective supplement to meals. The pod, cased in monosol, contains nutrients (like iron, omega 3, and folate) required by children under the age of five and caretaking mothers. Also created a distribution algorithm that creates an index that determines which states need the most funding to tackle malnutrition. Press mentions: ​​Infosys Science Foundation Nutrition Challenge (Collaborators: L. Jay, K. Dinakaran, P. Shekar, S. Rao)

Predictive Analytics for Barley Yields We analyzed the relationship between barley yields in the United States and climate change factors using open data and data analysis techniques and devised a solution using IoT and data science to increase yield. Developed a system for farmers to use data-backed decisions to maximize yields. The solution implements IoT which uses sensors and internet connectivity to gather and analyze data. (Collaborators: A. Parekh, K. McManus, J. Arevalo)