What it’s like being an Adobe intern in Bucharest

When Adrian-Stefan Ungureanu completed his PhD, he was looking for the right opportunity to utilize all his skills. Like many candidates, he was looking for an internship where he would be able to work on cutting-edge projects, so when he was given the chance to work on creating new Photoshop features, he jumped at the opportunity. Here’s how he spent his internship with the Adobe Romania team.

Why did you originally choose an internship at Adobe?

I chose the internship at Adobe thanks to the innovative projects that we would be engaged with, as well as the skills that I would sharpen in the process. In the past, I had used Photoshop for some time and became very fond of it. Being able to work around Photoshop-specific projects seemed like a dream come true. Ever since deep learning broke into the artistic sphere with features like Style Transfer, I’ve always wondered why Adobe hadn’t integrated this into Photoshop or other products. Well, imagine my surprise when I found out that I would be working with the team integrating features like these!

What are you studying now and how is it tied to your internship at Adobe?

This year marked the year of my graduation from the National University of Ireland, Galway, where I completed my PhD degree. I worked on using machine learning approaches to explore biometric recognition on smartphones. Returning home to Bucharest, I was looking for an opportunity to use the skills that I gained in research and computer vision.

What project have you been working on?

My project was around exploring automatic image colorization, where a neural network is tasked with making color predictions for each pixel in a grayscale image. What we do is take a normal colored image (RGB) and obtain the grayscale equivalent, so that the color information can be used to compute a distance between the network’s output and the expected output.

There are many things to consider here, from architecture to loss functions and training strategies, so it’s not an easy task, especially to reach the current state-of-the-art performance. Working on this project and discovering solutions with the team was the best part of my entire internship.

What are you most proud of accomplishing during your internship?

Becoming more comfortable with using the Pytorch framework for training machine learning algorithms. I’ve used Tensorflow and Keras previously, but more from a higher level. This internship gave me the opportunity to get my hands dirty and spend most of my time with a tool that’s used across the industry for computer vision applications.

Tell us about the Intern Summit. What did you enjoy most about it?

The virtual summit, which gave us the opportunity to hear from former interns (now employees) and current interns, was great. It was very insightful hearing about their experience, as well as leaders from across the organization. Working from home has made it difficult to interact with each other, so an event like this helped us connect.

Do you have any advice for someone who is thinking about applying to Adobe’s internship program?

Go for it! The experience has been great one and you will leave more knowledgeable than when you started. Expect to learn a lot—about the project and yourself! Every challenge is an opportunity and my team always provided me with support whenever I needed guidance or help.

Tell us about the recruitment process. What should a student expect to happen?

Be thorough about your background during the technical interview, because you will need to pull out information all the way from your first years of university. However, one shouldn’t be afraid of this process. Be proud of your achievements and describe your accomplishments, whatever they are—personal projects or competitions.

Considering your experience so far, how would you describe Adobe in three words?

Creativity, excellence, and supportiveness.