Search

In the realm of smart sensing with the Internet of Things, earable devices are empowered with the capability of multi-modality sensing and intelligence of context-aware computing, leading to its wide usage in Human Activity Recognition (HAR). Nonetheless, unlike the movements captured by Inertial Measurement Unit (IMU) sensors placed on the upper or lower body, those motion signals obtained from earable devices show significant changes in amplitudes and patterns, especially in the presence of dynamic and unpredictable head movements, posing a significant challenge for activity classification. In this work, we present EarDA, an adversarial-based domain adaptation system to extract the domain-independent features across different sensor locations. Moreover, while most deep learning methods commonly rely on training with substantial amounts of labeled data to offer good accuracy, the proposed scheme can release the potential usage of publicly available smartphone-based IMU datasets. Furthermore, we explore the feasibility of applying a filter-based data processing method to mitigate the impact of head movement. EarDA, the proposed system, enables more data-efficient and accurate activity sensing. It achieves an accuracy of 88.8% under HAR task, demonstrating a significant 43% improvement over methods without domain adaptation. This clearly showcases its effectiveness in mitigating domain gaps., Comment: accepted by 2024 IEEE Coupling of Sensing & Computing in AIoT Systems (CSCAIoT)

Cryogenic temperatures are the prerequisite for many advanced scientific applications and technologies. The accurate determination of temperature in this range and at the submicrometer scale is, however, nontrivial. This is due to the fact that temperature reading in cryogenic conditions can be inaccurate due to optically induced heating. Here, we present an ultralow power, optical thermometry technique that operates at cryogenic temperatures. The technique exploits the temperature dependent linewidth broadening measured by resonant photoluminescence of a two level system, a germanium vacancy color center in a nanodiamond host. The proposed technique achieves a relative sensitivity of 20% 1/K, at 5 K. This is higher than any other all optical nanothermometry method. Additionally, it achieves such sensitivities while employing excitation powers of just a few tens of nanowatts, several orders of magnitude lower than other traditional optical thermometry protocols. To showcase the performance of the method, we demonstrate its ability to accurately read out local differences in temperatures at various target locations of a custom-made microcircuit. Our work is a definite step towards the advancement of nanoscale optical thermometry at cryogenic temperatures.

All optical nanothermometry has become a powerful, noninvasive tool for measuring nanoscale temperatures in applications ranging from medicine to nanooptics and solid-state nanodevices. The key features of any candidate nanothermometer are sensitivity and resolution. Here, we demonstrate a real time, diamond based nanothermometry technique with sensitivity and resolution much larger than those of any existing all optical method. The distinct performance of our approach stems from two factors. First, temperature sensors nanodiamonds cohosting two Group IV colour centers engineered to emit spectrally separated Stokes and AntiStokes fluorescence signals under excitation by a single laser source. Second, a parallel detection scheme based on filtering optics and high sensitivity photon counters for fast readout. We demonstrate the performance of our method by monitoring temporal changes in the local temperature of a microcircuit and a MoTe2 field effect transistor. Our work lays the foundation for time resolved temperature monitoring and mapping of micro, nanoscale devices such as microfluidic channels, nanophotonic circuits, and nanoelectronic devices, as well as complex biological environments such as tissues and cells.

Hexagonal boron nitride (hBN) is gaining considerable attention as a solid-state host of quantum emitters from the ultraviolet to the near infrared spectral ranges. However, atomic structures of most of the emitters are speculative or unknown, and emitter fabrication methods typically suffer from poor reproducibility, spatial accuracy, or spectral specificity. Here, we present a robust, deterministic electron beam technique for site-specific fabrication of blue quantum emitters with a zero-phonon line at 436 nm (2.8 eV). We show that the emission intensity is proportional to electron dose and that the efficacy of the fabrication method correlates with a defect emission at 305 nm (4.1 eV). We attribute blue emitter generation to fragmentation of carbon clusters by electron impact and show that the robustness and universality of the emitter fabrication technique are enhanced by a pre-irradiation annealing treatment. Our results provide important insights into photophysical properties and structure of defects in hBN and a framework for deterministic fabrication of quantum emitters in hBN.

Integration of solid state quantum emitters into nanophotonic circuits is a critical step towards fully on-chip quantum photonic based technologies. Among potential materials platforms, quantum emitters in hexagonal boron nitride have emerged over the last years as viable candidate. While the fundamental physical properties have been intensively studied over the last years, only few works have focused on the emitter integration into photonic resonators. Yet, for a potential quantum photonic material platform, the integration with nanophotonic cavities is an important cornerstone, as it enables the deliberate tuning of the spontaneous emission and the improved readout of distinct transitions for that quantum emitter. In this work, we demonstrate the resonant tuning of an integrated monolithic hBN quantum emitter in a photonic crystal cavity through gas condensation at cryogenic temperature. We resonantly coupled the zero phonon line of the emitter to a cavity mode and demonstrate emission enhancement and lifetime reduction, with an estimation for the Purcell factor of ~ 15.

Fluorescent nanoparticles are widely utilized in a large range of nanoscale imaging and sensing applications. While ultra-small nanoparticles (size <10 nm) are highly desirable, at this size range their photostability can be compromised due to effects such as intensity fluctuation and spectral diffusion caused by interaction with surface states. In this letter, we demonstrate a facile, bottom-up technique for the fabrication of sub-10-nm hBN nanoparticles hosting photostable bright emitters via a catalyst-free hydrothermal reaction between boric acid and melamine. We also implement a simple stabilization protocol that significantly reduces intensity fluctuation by ~85% and narrows the emission linewidth by ~14% by employing a common sol-gel silica coating process. Our study advances a promising strategy for the scalable, bottom-up synthesis of high-quality quantum emitters in hBN nanoparticles.

Nanoscale optical thermometry is a promising non-contact route for measuring local temperature with both high sensitivity and spatial resolution. In this work, we present a deterministic optical thermometry technique based on quantum emitters in nanoscale hexagonal boron-nitride. We show that these nanothermometers exhibit better performance than that of homologous, all-optical nanothermometers both in sensitivity and range of working temperature. We demonstrate their effectiveness as nanothermometers by monitoring the local temperature at specific locations in a variety of custom-built micro-circuits. This work opens new avenues for nanoscale temperature measurements and heat flow studies in miniaturized, integrated devices.

Benefiting from the great success of deep learning in computer vision, CNN-based object detection methods have drawn significant attentions. Various frameworks have been proposed which show awesome and robust performance for a large range of datasets. However, for building detection in remote sensing images, buildings always pose a diversity of orientations which makes it a challenge for the application of off-the-shelf methods to building detection. In this work, we aim to integrate orientation regression into the popular axis-aligned bounding-box detection method to tackle this problem. To adapt the axis-aligned bounding boxes to arbitrarily orientated ones, we also develop an algorithm to estimate the Intersection over Union (IoU) overlap between any two arbitrarily oriented boxes which is convenient to implement in Graphics Processing Unit (GPU) for accelerating computation. The proposed method utilizes CNN for both robust feature extraction and rotated bounding box regression. We present our modelin an end-to-end fashion making it easy to train. The model is formulated and trained to predict orientation, location and extent simultaneously obtaining tighter bounding box and hence, higher mean average precision (mAP). Experiments on remote sensing images of different scales shows a promising performance over the conventional one., Comment: Presented in DICTA2018; Further Improved including extra experiments