Mary Davids
LSAMP Scholar
Major: Political Science and Computer Science
Mentor: Lantao Liu
The above research constitutes of the application of technology to reducing the cases of sexual misconduct on the Indiana University campus. Considering the record-high rates of sexual misconduct in the past year, the goal is to allow the intervention of intelligent robots in situations of sexual misconduct in public spaces.
My research revolved around the functions of Unmanned Autonomous Vehicles (drones) in assessing dangerous scenarios and differentiating the aggressor from the victims using deep learning in image recognition technology. I experimented with the creation of an image recognition model that processed a wide range of inputs; starting with an established MNIST dataset, then progressing to a dataset with binary classes (jpg) and finally an unstructured dataset (png) that I had to preprocess. Further, I experimented with video (mp4) inputs considering the drone would be using a live stream essentially made of a series of images. At the moment, the model is capable of taking image and video inputs when adjusted accordingly with a 92% prediction accuracy. When further improved, such model can be trained to potentially identify a victim from an aggressor in an instance of sexual misconduct. The presentation also includes a map and an execution video.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. IN LSAMP is supported by NSF Award Number: HRD-1618408, 2016-2022.
Leave a Reply