Ensemble AI unlocks hidden tree crown structures in dense forests
Peer-Reviewed Publication
Updates every hour. Last Updated: 12-Nov-2025 23:11 ET (13-Nov-2025 04:11 GMT/UTC)
A research team has developed a novel approach combining refined spatial competition mapping with powerful ensemble learning to predict maximum crown width height (HMCW) in Chinese fir with unprecedented accuracy.
A research team has developed PlantCaFo, an advanced few-shot plant disease recognition model powered by foundation models, capable of achieving high accuracy with only a handful of training images.
A research team has developed a cost-effective method to measure wheat plant height in unprecedented detail using drone-based cross-circling oblique (CCO) imaging and 3D canopy modeling.
This study initiatively developed the phosphor-free LED lamp based on high-efficiency yellow LED, and explored the photo-biological effects of yellow-green spectrum. Analyzing multiple ocular physiological parameters, cortical functional connectivity, and melatonin secretions at various moments, the authors demonstrated the positive photo-biological effects of yellow-green-rich phosphor-free LED lamp. This study revealed the spectral regulation mechanism and the intrinsic mathematic model of human physiological response, applicable in the fields of health-oriented lighting and displays.
Background
As the market demands change, SMEs (small and medium-sized enterprises) have long faced many design issues, including high costs, lengthy cycles, and insufficient innovation. These issues are especially noticeable in the domain of cosmetic packaging design.
Objective
To explore innovative product family modeling methods and configuration design processes to improve the efficiency of enterprise cosmetic packaging design and develop the design for mass customization.
Methods
To accomplish this objective, the basic-element theory has been introduced and applied to the design and development system of the product family.
Results
By examining the mapping relationships between the demand domain, functional domain, technology domain, and structure domain, four interrelated models have been developed, including the demand model, functional model, technology model, and structure model. Together, these models form the mechanism and methodology of product family modeling, specifically for cosmetic packaging design. Through an analysis of a case study on men’s cosmetic packaging design, the feasibility of the proposed product family modeling technology has been demonstrated in terms of customized cosmetic packaging design, and the design efficiency has been enhanced.
Conclusion
The product family modeling technology employs a formalized element as a module configuration design language, permeating throughout the entire development cycle of cosmetic packaging design, thus facilitating a structured and modularized configuration design process for the product family system. The application of the basic-element principle in product family modeling technology contributes to the enrichment of the research field surrounding cosmetic packaging product family configuration design, while also providing valuable methods and references for enterprises aiming to elevate the efficiency of cosmetic packaging design for the mass customization product model.
UCLA and Broadcom researchers introduced the design and fabrication of a broadband, polarization-insensitive unidirectional imager operating in the visible spectrum. This device was fabricated through wafer-scale lithography on high-purity fused silica, offering high optical transparency, thermal stability and ultra-low loss. This work marks the first wafer-scale multilayer diffractive processor designed for visible imaging, enabling scalable, high-throughput production for advanced applications such as intelligent imaging, optical sensing, privacy protection, and all-optical information processing.
A research team led by Professor Sung Hoon Kim from the School of Electrical and Computer Engineering at Korea University has developed the world’s first alginate-based microrobot that can be tracked using Magnetic Particle Imaging (MPI). The newly developed system marks a major milestone in microrobotics by enabling real-time localization, selective thermal therapy, and cell delivery—all powered and controlled through a single magnetic actuation system, completely independent of conventional medical imaging devices such as cameras, CT, or X-ray machines.
Biochar, hailed as a climate change mitigation tool for its ability to reduce greenhouse gas emissions, may not be as effective as once thought when microplastics are in the picture. A new study investigates how polyethylene microplastics interact with aged biochar to affect carbon and nitrogen stability in Fluvic Cambisol soil, offering surprising results.
The development of efficient electrocatalysts for saline water oxidation (SWO) is imperative for advancing seawater splitting technology to produce green hydrogen. In this work, a NiFe-Co2(OH)3Cl material was developed as pre-electrocatalyst for both active and stable SWO. During catalysis, the hydroxyloride was converted to oxyhydroxide with 35.4% enlarged ECSA, leading to the enhanced intrinsic activity (300 mV@10 mA cm-2, 49.9 mV decade-1). Moreover, the electrolyte Cl- would be incorporated to the catalyst lattice, thus improving the corrosion-resistance of the material, resulting in the high electrocatalytic stability for 100 h.