Deep-learning-based gas identification by time-variant illumination of a single micro-LED-embedded gas sensor.

Electronic nose (e-nose) technology for selectively identifying a target gas through chemoresistive sensors has gained much attention for various applications, such as smart factory and personal health monitoring. To overcome the cross-reactivity problem of chemoresistive sensors to various gas species, herein, we propose a novel sensing strategy based on a single micro-LED (µLED)-embedded photoactivated (µLP) gas sensor, utilizing the time-variant illumination for identifying the species and concentrations of various target gases. A fast-changing pseudorandom voltage input is applied to the µLED to generate forced transient sensor responses. A deep neural network is employed to analyze the obtained complex transient signals for gas detection and concentration estimation. The proposed sensor system achieves high classification (~96.99%) and quantification (mean absolute percentage error ~ 31.99%) accuracies for various toxic gases (methanol, ethanol, acetone, and nitrogen dioxide) with a single gas sensor consuming 0.53 mW. The proposed method may significantly improve the efficiency of e-nose technology in terms of cost, space, and power consumption.

Ultra-Low-Power E-Nose System Based on Multi-Micro-LED-Integrated, Nanostructured Gas Sensors and Deep Learning.

As interests in air quality monitoring related to environmental pollution and industrial safety increase, demands for gas sensors are rapidly increasing. Among various gas sensor types, the semiconductor metal oxide (SMO)-type sensor has advantages of high sensitivity, low cost, mass production, and small size but suffers from poor selectivity. To solve this problem, electronic nose (e-nose) systems using a gas sensor array and pattern recognition are widely used. However, as the number of sensors in the e-nose system increases, total power consumption also increases. In this study, an ultra-low-power e-nose system was developed using ultraviolet (UV) micro-LED (µLED) gas sensors and a convolutional neural network (CNN). A monolithic photoactivated gas sensor was developed by depositing a nanocolumnar In2O3 film coated with plasmonic metal nanoparticles (NPs) directly on the µLED. The e-nose system consists of two different µLED sensors with silver and gold NP coating, and the total power consumption was measured as 0.38 mW, which is one-hundredth of the conventional heater-based e-nose system. Responses to various target gases measured by multi-µLED gas sensors were analyzed by pattern recognition and used as the training data for the CNN algorithm. As a result, a real-time, highly selective e-nose system with a gas classification accuracy of 99.32% and a gas concentration regression error (mean absolute) of 13.82% for five different gases (air, ethanol, NO2, acetone, methanol) was developed. The µLED-based e-nose system can be stably battery-driven for a long period and is expected to be widely used in environmental internet of things (IoT) applications.

Artificial Olfactory Neuron for an In-Sensor Neuromorphic Nose.

A neuromorphic module of an electronic nose (E-nose) is demonstrated by hybridizing a chemoresistive gas sensor made of a semiconductor metal oxide (SMO) and a single transistor neuron (1T-neuron) made of a metal-oxide-semiconductor field-effect transistor (MOSFET). By mimicking a biological olfactory neuron, it simultaneously detects a gas and encoded spike signals for in-sensor neuromorphic functioning. It identifies an odor source by analyzing the complicated mixed signals using a spiking neural network (SNN). The proposed E-nose does not require conversion circuits, which are essential for processing the sensory signals between the sensor array and processors in the conventional bulky E-nose. In addition, they do not have to include a central processing unit (CPU) and memory, which are required for von Neumann computing. The spike transmission of the biological olfactory system, which is known to be the main factor for reducing power consumption, is realized with the SNN for power savings compared to the conventional E-nose with a deep neural network (DNN). Therefore, the proposed neuromorphic E-nose is promising for application to Internet of Things (IoT), which demands a highly scalable and energy-efficient system. As a practical example, it is employed as an electronic sommelier by classifying different types of wines.

High Accuracy Real-Time Multi-Gas Identification by a Batch-Uniform Gas Sensor Array and Deep Learning Algorithm.

Semiconductor metal oxide (SMO) gas sensors are attracting great attention as next-generation environmental monitoring sensors. However, there are limitations to the actual application of SMO gas sensors due to their low selectivity. Although the electronic nose (E-nose) systems based on a sensor array are regarded as a solution for the selectivity issue, poor accuracy caused by the nonuniformity of the fabricated gas sensors and difficulty of real-time gas detection have yet to be resolved. In this study, these problems have been solved by fabricating uniform gas sensor arrays and applying the deep learning algorithm to the data from the sensor arrays. Nanocolumnar films of metal oxides (SnO2, In2O3, WO3, and CuO) with a high batch uniformity deposited through glancing angle deposition were used as the sensing materials. The convolutional neural network (CNN) using the input data as a matrix form was adopted as a learning algorithm, which could conduct pattern recognition of the sensor responses. Finally, real-time selective gas detection for CO, NH3, NO2, CH4, and acetone (C3H6O) gas was achieved (minimum response time of 1, 8, 5, 19, and 2 s, respectively) with an accuracy of 98% by applying preprocessed response data to the CNN.

SpherePHD: Applying CNNs on 360 ° Images With Non-Euclidean Spherical PolyHeDron Representation.

Omni-directional images are becoming more prevalent for understanding the scene of all directions around a camera, as they provide a much wider field-of-view (FoV) compared to conventional images. In this work, we present a novel approach to represent omni-directional images and suggest how to apply CNNs on the proposed image representation. The proposed image representation method utilizes a spherical polyhedron to reduce distortion introduced inevitably when sampling pixels on a non-Euclidean spherical surface around the camera center. To apply convolution operation on our representation of images, we stack the neighboring pixels on top of each pixel and multiply with trainable parameters. This approach enables us to apply the same CNN architectures used in conventional euclidean 2D images on our proposed method in a straightforward manner. Compared to the previous work, we additionally compare different designs of kernels that can be applied to our proposed method. We also show that our method outperforms in monocular depth estimation task compared to other state-of-the-art representation methods of omni-directional images. In addition, we propose a novel method to fit bounding ellipses of arbitrary orientation using object detection networks and apply it to an omni-directional real-world human detection dataset.

Binodal, wireless epidermal electronic systems with in-sensor analytics for neonatal intensive care.

Existing vital sign monitoring systems in the neonatal intensive care unit (NICU) require multiple wires connected to rigid sensors with strongly adherent interfaces to the skin. We introduce a pair of ultrathin, soft, skin-like electronic devices whose coordinated, wireless operation reproduces the functionality of these traditional technologies but bypasses their intrinsic limitations. The enabling advances in engineering science include designs that support wireless, battery-free operation; real-time, in-sensor data analytics; time-synchronized, continuous data streaming; soft mechanics and gentle adhesive interfaces to the skin; and compatibility with visual inspection and with medical imaging techniques used in the NICU. Preliminary studies on neonates admitted to operating NICUs demonstrate performance comparable to the most advanced clinical-standard monitoring systems.