Deep learning transforms how machines perceive and interpret information by stacking layers of artificial neurons to automatically extract increasingly abstract features. Its evolution has accelerated dramatically in recent years, progressing from convolutional neural networks to today’s state-of-the-art transformer architectures that power advanced reasoning systems capable of understanding multiple modalities simultaneously.
At AIRLab, we leverage cutting-edge neural architectures for computer vision, large language models, and multimodal systems that combine visual, textual and other forms of perceptual understanding, as well as generative applications. Our investigations advance the theoretical foundations and practical implementations of neural networks, enhancing accuracy, robustness, privacy and efficiency across biomedical data analysis, industrial applications, and intelligent environments.
Contacts
Anomaly Detection
The goal of fully automating anomaly detection across all industrial quality control processes through videos or images remains a technological...
Artificial Life
"From so simple a beginning endless forms most beautiful and most wonderful have been, and are being, evolved."On the Origin...
Generative Deep Learning
In regulated, highly variable settings, the supply of real-world data no longer grows at the pace demanded by modern models:...
Multimodal Learning
Multimodal Learning: Bridging the Senses of AI In our everyday lives, we effortlessly weave together vision, sound, touch, language, and...
Survival Analysis
Survival analysis is a branch of statistics focused on modeling time-to-event data, playing a critical role in healthcare by informing...