Back to projects
Nov 10, 2019
1 min read

Samsung PRISM

Contributing to the Samsung Prism initiative

Worked in a team to create a dataset for detection of ambient temperature on SAMSUNG Mobile devices without the use of specialized hardware

Data was collected via a native Android app using different Samsung phones and aggregated on a server where it was used to train and deploy an on-device predictive model with high accuracy. This model was used to predict the temperature of a device on phones that did not have a built-in temperature sensor

The model performed well and from what I was told, is in use on modern Samsung Galaxy devices