MBZUAI's Samuel Horváth presented a new framework called Maestro at ICML 2024 for efficiently training machine learning models in federated settings. Maestro identifies and removes redundant components of a model through trainable decomposition to increase efficiency on edge devices. The approach decomposes layers into low-dimensional approximations, discarding unused aspects to reduce model size. Why it matters: This research addresses the challenge of running complex models on resource-constrained devices, crucial for expanding AI applications while preserving data privacy.
A KAUST scientist led a global call for climate solutions, published simultaneously by 14 academic journals and released at COP29. The publication, prepared by 18 scientists, urges international governments to deploy microbial 'vaccines' against climate change. Six simple 'vaccine' examples are outlined, including carbon sequestration boosters and methane busters. Why it matters: This coordinated effort highlights the urgency of addressing climate change and KAUST's leading role in microbial solutions.
Eliseo Ferrante from NYU Abu Dhabi presented work on increasing the controllability of swarm robotics systems. The research covers microscopic control via implicit intelligent leaders and macroscopic control via automated generation of swarm behaviors. Grammatical evolution and generative AI methods are used to produce collective behaviors aligned with human specifications. Why it matters: This research enhances the applicability of swarm robotics in real-world scenarios by improving control and coordination, potentially impacting industries like logistics, environmental monitoring, and disaster response in the region.
This paper introduces Diffusion-BBO, a new online black-box optimization (BBO) framework that uses a conditional diffusion model as an inverse surrogate model. The framework employs an Uncertainty-aware Exploration (UaE) acquisition function to propose scores in the objective space for conditional sampling. The approach is shown theoretically to achieve a near-optimal solution and empirically outperforms existing online BBO baselines across 6 scientific discovery tasks.
This paper presents a decentralized multi-agent unmanned aerial system designed for search, pickup, and relocation of objects. The system integrates multi-agent aerial exploration, object detection/tracking, and aerial gripping. The decentralized system uses global state estimation, reactive collision avoidance, and sweep planning for exploration. Why it matters: The system's successful deployment in demonstrations and competitions like MBZIRC highlights the potential of integrated robotic solutions for complex tasks such as search and rescue in the region.
MBZUAI's AI Quorum launched its second workshop, "Building Ecosystems for AI at Scale," focusing on AI scalability and business applications. The first CASL workshop aims to define steps for organizations to become self-sufficient with AI and explore new use cases. Speakers include MBZUAI faculty and researchers from CMU, Stanford, KAUST, UC Berkeley, and Google. Why it matters: The workshop highlights the UAE's growing role in fostering AI innovation and bridging the gap between academic research and industry applications in the region.