Peter Baldry

Over the 2018/2019 summer I interned at Grabba, who build software and hardware solutions for data capture, data verification, transaction processing and more. My internship was about biometric identification - more specifically facial recognition. I successfully developed a model to identify individuals from imagery of their face sourced from CCTV or security camera footage.

The first half of the internship was about extracting numerical features of the face from the images. I used several computer vision algorithms and numerical methods in Python to extract valuable features. The second half of the internship was focused on the matching algorithm. Here, I implemented several machine learning algorithms to match the individual in the target images to a known individual in a database. This part included training, validation and testing of the model.

In short, I learnt a lot! It was great to apply the skills I had learnt from my degree in an industry project as well as experience the industry work environment. I had an amazing supervisor who helped keep my project on track and gave me lots of advice about working as a data scientist in industry. The rest of the team at Grabba were very inviting and were always willing to assist me. After the internship, I was fortunate enough to get part time work with Grabba, where I’m currently working on another exciting project (flexibly around my university studies). I highly recommend applying as you will be sure to learn lots, make connections and gain invaluable experience.