Earlier this year, I had the incredible opportunity to spend four months in Berlin through a research stay supported by the “Ajut Complementari per a la Mobilitat del Personal Investigador en Formació IDUdG (IF MOB 2023)” grant. As part of my PhD at VICOROB, I joined the Explainable AI Lab at the Fraunhofer Heinrich-Hertz-Institute (HHI), where I worked on multi-modal explainability for dermoscopic skin lesion diagnosis.
From the moment I arrived, Berlin’s unique blend of history, creativity, and science left a lasting impression. Each day felt like a new layer unfolding, academically challenging, culturally rich, and personally empowering.
Back at VICOROB, within the iToBoS project, my research focused on developing machine learning models for the early diagnosis of skin cancer. Our aim was to create models that could eventually be used in clinical settings, supporting dermatologists with faster and more accurate decisions. But clinical deployment isn’t just about accuracy, it’s also about trust. Models need to be not only effective but also transparent and safe. This challenge led me to HHI, where I focused on building multi-modal vision-and-language models that could explain their predictions at a professional level, rather than just output a diagnosis.
The first step was to explore and benchmark existing models designed for similar tasks, understanding where they performed well and where they fell short. With those insights, I began adapting and improving them to better serve the needs of dermatological applications. Throughout this process, I had the privilege of collaborating closely with other PhD students and researchers at HHI. Their expertise in vision-language models and Explainable AI, as well as their openness to share knowledge, made a huge impact on the development of my work.
Beyond the technical growth, the experience of working in a new environment was refreshing and energizing. I noticed the differences, and appreciated the commonalities, between VICOROB and HHI, which helped me grow not only as a researcher but also as a team member in a broader scientific community. And there was something quietly magical about the routine itself: each morning on my way to the office in Charlottenburg, I passed the Brandenburg Gate. No matter how many times I walked by, it never failed to inspire me with its grandeur and historical presence.
Outside of work, Berlin offered so much more than I expected. The city’s energy, diversity, and openness created space for learning far beyond the lab. As a woman in tech, I found particular inspiration in the local community. Through the BLISS e.V. initiative, I joined career development events designed for women in machine learning and computer vision. There, I met a number of accomplished women whose stories of resilience and success deeply resonated with me.Another standout moment was attending “Redefine Possible,” an event hosted by Google Developers Berlin to celebrate International Women’s Day. Surrounded by driven young women from across Germany, we reflected on how far the field has come, and how much more we can do to foster equity and inclusion in science and tech. These experiences reminded me of the importance of community, visibility, and mentorship.
Overall, my stay in Berlin was an incredibly enriching experience, both academically and personally. Working at Fraunhofer HHI deepened my research and helped me grow as a scientist, while the diverse events and inspiring people I met along the way expanded my vision of what’s possible in my field. I returned home not only with valuable progress on my PhD, but also with renewed confidence, fresh ideas, and a stronger sense of belonging within the global research community. This stay has left a lasting impact on me, and I’m deeply grateful for the opportunity.
