We are focusing on Continual Learning in AI, which involves training machine learning models to learn continuously from new data while maintaining previously learned information. This approach prevents 'catastrophic forgetting' and ensures our AI systems adapt and retain knowledge effectively over time.
We are exploring innovative methods to enhance AI, making it more adaptable and efficient. Our approach focuses on teaching AI to learn in ways similar to humans, better preparing it for a variety of real-world challenges. Our goal is to develop AI that is more capable of handling diverse tasks and learning effectively from different experiences.
We are advancing the field of Deep Generative Models, aiming to improve how AI creates and understands visuals. Our work focuses on making better images, animating pictures, and building 3D models with AI. Our goal is to enhance AI's ability to perform complex visual tasks more easily.
We are working on Multi-Modal AI to teach systems how to handle and make sense of various kinds of data at once. Our goal is to build smarter AI that can make good predictions and spot things that don't follow the usual patterns, making it better at solving real-world problems.