No abstract

The three-dimensional (3D) vectorial nature of electromagnetic waves of light has not only played a fundamental role in science but also driven disruptive applications in optical display, microscopy, and manipulation. However, conventional optical holography can address only the amplitude and phase information of an optical beam, leaving the 3D vectorial feature of light completely inaccessible. We demonstrate 3D vectorial holography where an arbitrary 3D vectorial field distribution on a wavefront can be precisely reconstructed using the machine learning inverse design based on multilayer perceptron artificial neural networks. This 3D vectorial holography allows the lensless reconstruction of a 3D vectorial holographic image with an ultrawide viewing angle of 94° and a high diffraction efficiency of 78%, necessary for floating displays. The results provide an artificial intelligence–enabled holographic paradigm for harnessing the vectorial nature of light, enabling new machine learning strategies for holographic 3D vectorial fields multiplexing in display and encryption.

Smartphones, wearables, and Internet of Things (IoT) devices produce a wealth of data that cannot be accumulated in a centralized repository for learning supervised models due to privacy, bandwidth limitations, and the prohibitive cost of annotations. Federated learning provides a compelling framework for learning models from decentralized data, but conventionally, it assumes the availability of labeled samples, whereas on-device data are generally either unlabeled or cannot be annotated readily through user interaction. To address these issues, we propose a self-supervised approach termed scalogramsignal correspondence learning based on wavelet transform to learn useful representations from unlabeled sensor inputs, such as electroencephalography, blood volume pulse, accelerometer, and WiFi channel state information. Our auxiliary task requires a deep temporal neural network to determine if a given pair of a signal and its complementary viewpoint (i.e., a scalogram generated with a wavelet transform) align with each other or not through optimizing a contrastive objective. We extensively assess the quality of learned features with our multi-view strategy on diverse public datasets, achieving strong performance in all domains. We demonstrate the effectiveness of representations learned from an unlabeled input collection on downstream tasks with training a linear classifier over pretrained network, usefulness in low-data regime, transfer learning, and cross-validation. Our methodology achieves competitive performance with fullysupervised networks, and it outperforms pre-training with autoencoders in both central and federated contexts. Notably, it improves the generalization in a semi-supervised setting as it reduces the volume of labeled data required through leveraging self-supervised learning.

In the last four decades several methods have been used to model occupants' presence and actions (OPA) in buildings according to different purposes, available computational power, and technical solutions. This study reviews approaches, methods and key findings related to OPA modeling in buildings. An extensive database of related research documents is systematically constructed, and, using bibliometric analysis techniques, the scientific production and landscape are described. The initial literature screening identified more than 750 studies, out of which 278 publications were selected. They provide an overarching view of the development of OPA modeling methods. The research field has evolved from longitudinal collaborative efforts since the late 1970s and, so far, covers diverse building typologies mostly concentrated in a few climate zones. The modeling approaches in the selected literature are grouped into three categories (rule-based models, stochastic OPA modeling, and data-driven methods) for modeling occupancy-related target functions and a set of occupants' actions (window, solar shading, electric lighting, thermostat adjustment, clothing adjustment and appliance use). The explanatory modeling is conventionally based on the model-based paradigm where occupant behavior is assumed to be stochastic, while the Revised Manuscript with No Changes MarkedClick here to view linked References data-driven paradigm has found wide applications for the predictive modeling of OPA, applicable to control systems. The lack of established standard evaluation protocols was identified as a scientifically important yet rarely addressed research question. In addition, machine learning and deep learning are emerging in recent years as promising methods to address OPA modeling in real-world applications.

No abstract