Alessio Quercia
CS PhD Candidate @ RWTH Aachen University & FZJ | ex IBM Research Zurich, WSense
Biography
I am a Ph.D. Candidate in Computer Science at RWTH Aachen University and at the Machine Learning and Data Analytics Institute in Forschungszentrum Jülich. Previously, I was a Ph.D. Candidate in Computer Science at Sapienza University of Rome and at WSense, and a Research Intern in the Smart Systems Integration group at IBM Research Zurich. I obtained my Master’s Degree in Computer Science from University of Milan and my Bachelor’s Degree in Computer Science from Sapienza University of Rome.
I am mainly interested in Deep Learning and Computer Vision applications to real-world problems. My research interests include Data Efficient Learning, Transfer Learning, Parameter-Efficient Fine-Tuning (PEFT) methods, Multi-Task and Multi-Modal Learning.
I have experience with the reviewing process of top AI conferences such as NeurIPS, ICLR, ICML, ECCV, ICCV, CVPR.
I am always open to new collaborations. If you are interested in collaborating with me or in my detailed Curriculum Vitae, please get in contact with me using any of the specified contact options.
-
Italian
Native
-
English
Fluent
-
French
Professional
-
German
Beginner
- Data Efficient Learning
- Transfer Learning
- Parameter-Efficient Fine-Tuning (PEFT)
- Multi-Task Learning
- Multi-Modal Learning
- In-Context Learning
-
PhD in Computer Science, Current
-
PhD in Computer Science, 2021
-
MSc in Computer Science, 2020
110/110
-
BSc in Computer Science, 2017
110/110
Experience
Responsibilities
- Research on Data and Knowledge Transfer Efficiency, using Data Pruning, Transfer Learning, Multi-Task Learning, Parameter-Efficient Fine-Tuning and Uncertainty Quantification methods.
- Publication in top (A, A*) conferences (ICDM, ECCV, WACV) of state-of-the-art methods in the areas of Data Efficiency, Transfer Learning and Multi-Task Learning (Monocular Depth Estimation, Semantic Segmentation and Image Classification).
- Design and development of state-of-the-art Transfer Learning methods for medical imaging.
- Experience with training large vision and language models on multi-node and multi-gpu HPC systems (JURECA, JUWELS).
- Experience with ResNets, ViTs, SAM, DINOv2, DepthAnything, DiTs, LLaMA2.
- Experience with Image Classification, Super-Resolution, Monocular Depth Estimation, Semantic Segmentation, Image Generation, Text Generation.
Responsibilities
- Research activity on risk forecasting in aquaculture using machine learning and deep learning (unsupervised learning techniques) for the EASME/EMFF SeaStar EUCORE project.
- Analyzing data from in-fish sensors in order to detect and predict their health issues.
Responsibilities
- Research on Epileptic Seizure Detection from non-EEG signals (EMG, ECG, EDA, …), collected from non-invasive devices.
- Research on Epileptic Seizure Prediction from EEG signals using deep learning.
- Development of a pipeline for Seizure Prediction from EEG signals using both unsupervised and supervised learning techniques.
Responsibilities
- Teaching Computer Science basic topics (basic programming, spreadsheets, databases) to up to 60 undergraduates in laboratory classes.
- Supporting students via email for any help.
- Examining students during the final course examinations.