Fingerprint Reader

Fingerprint Reader

 I decided that I would share with you not only what represents the range of my scientific interests today, but also my developments from the past.

Since childhood, I spent a lot of time in scientific research, read a lot and was in the state of a researcher. Having intentions to transform the world, to create benefits in it, I was up to a certain extent to creativity. Since childhood, I have loved not only working with code, but also making prototypes with my own hands. With this publication, perhaps, I will open a series of posts about my  inventions that I managed to bring to life.

Today I will tell you about one of my most stimulating STEM experiences. This was participation in a neural network programming project led by MIT graduate Ryan M. Sander. The goal was to build a working neural network in Python using the Keras API to compare scanned and saved fingerprints.

To test the software, I designed a device that read a fingerprint and, if the comparison was correct, displayed the text “Congratulations” on a small screen measuring 96×64 inches.

At that time, I had nothing with me but a soldering iron and a desire to create something wonderful. A deep sense of satisfaction that I see the line with the inscription “Congratulations!” flashing twice, no matter how insignificant it may seem, will remain in my memory forever.

This project was my first deep dive into machine learning and how neural networks analyze information. I remember thinking how confusing all these connections were, but the time-consuming research and problem solving was worth it. This sense of satisfaction prompted me to continue studying neural networks and image recognition. I want to have a hand in creating AGI and for this reason I would like to learn more about machine learning and the extraordinary ability of neural networks to both reproduce and surpass human thought processes.

Leave a Reply

Your email address will not be published. Required fields are marked *