果冻传媒

Our research activities focus on the design and implementation of decentralized AI systems that enable learning and analytics on distributed devices of everyday use. We develop advanced algorithms for federated learning, allowing models to be trained on data from multiple sources without compromising privacy. Additionally, we explore self-supervised learning techniques that enable AI systems to learn from unlabeled data, reducing the reliance on costly annotated datasets. In the realm of Human-Centric AI, we concentrate on developing AI systems that are transparent, explainable, and aligned with human values. We investigate methods for incorporating human feedback and preferences into the learning process, ensuring that AI systems remain accountable and beneficial to society.

Our mission at DARL is to push the boundaries of AI research, creating decentralized systems that are robust, adaptable, and capable of continuous learning, with a keen eye on their applications in Healthcare and High-Tech Industries. By harnessing the power of collective intelligence and embracing the principles of decentralization, we strive to unlock the full potential of AI and drive transformative advancements in these critical sectors.