Womb Wise: An Academic Adventure

In the past 2 years, I was working on some super interesting activities with my super amazing employer (Giulia).

This was her post doc project and I did some deep neural network development, something I had only done during my bachelor’s degree experimental thesis. I’m grateful for the opportunity Giulia gave me :)

Giulia asked me to implement an automatic machine-learning-driven pipeline to detect events and classify them in 2 groups of videos.

• The first group of videos were recordings of female faces during various exercises to simulate yawning and normal opening and closing of the mouth.
• The second group of videos were about the fetuses inside the wombs of those females.

I know. First reaction, shocked.

I implemented the LSTM architecture using PyTorch and tried many different enhancing tactics. I have a public repo for this project at git.kanopo.dev/womb-wise/womb-wise-public.

• Link to the publication
• link to download pdf from this website

I really enjoyed this experience and I hope to work in this field again.

Bye Bye