Scientists have developed a pioneering framework that translates human brain activity into editable visual imagery, opening up new possibilities for creative design and human–computer interaction. Named DreamConnect, the system employs a dual-stream diffusion model to directly interpret functional magnetic resonance imaging (fMRI) signals and refine them with natural language instructions. By progressively aligning brain activity with user-directed prompts, the method allows for manipulation of imagined scenes—such as transforming a mental picture of a lake into a vivid sunset. This breakthrough demonstrates the potential of brain-to-image technologies to actively shape human “dreams,” suggesting future applications in design, therapy, and communication.

Unlocking deep oil reservoirs just got easier! Scientists have developed a groundbreaking nanographite system that boosts oil recovery in extreme conditions. Read on to discover how this innovative solution overcomes high-temperature and high-salinity challenges, offering a game-changing approach for enhanced oil extraction.

Origami-inspired deployable structures are promising owing to their compact storage and efficient deployment. Recently, a team of researchers from Pusan National University have made an innovative breakthrough in this field. They present a multi-resin dispensing process for creating monolithic fiber-reinforced polymer structures with selective rigidity and flexibility, enabling origami-inspired deployable robotics applications with precise mechanical property control and mass production potential.

Embodied intelligence applications, such as autonomous robotics and smart transportation systems, require efficient coordination of multiple agents in dynamic environments. A critical challenge in this domain is the multi-agent pathfinding (MAPF) problem, which ensures that agents can navigate conflict-free while optimizing their paths. Conflict-based search (CBS) is a well-established two-level solver for the MAPF problem. However, as the scale of the problem expands, the computation time becomes a significant challenge for the implementation of CBS. Previous optimizations have mainly focused on reducing the number of nodes explored by the high-level or low-level solver. This paper takes a different perspective by proposing a parallel version of CBS, namely GPU-accelerated conflict-based search (GACBS), which significantly exploits the parallel computing capabilities of GPU. GACBS employs a task coordination framework to enable collaboration between the high-level and low-level solvers with lightweight synchronous operations. Moreover, GACBS leverages a parallel low-level solver, called GATSA, to efficiently find the shortest path for a single agent under constraints. Experimental results show that the proposed GACBS significantly outperforms CPU-based CBS, with the maximum speedup ratio reaching over 46.