Keywords: MR, Interaction, Prototyping, Deep learning
We present Knock the Reality, an interaction technique for virtual interface registration in mixed reality (MR). When a user knocks on a physical object, our technique identifies the object based on a knocking sound and registers a customizable virtual interface onto the object. Unlike computer vision-based methods, our approach does not require continuously processing image information. Instead, we utilize audio features which are less computationally expensive. This work presents our implementation and demonstrates an interaction scenario where a user works in MR. Overall, our method offers a simple and intuitive way to register MR interfaces.
Keywords: NIRS, Imaging, Prototyping, Deep learning
Non-invasive inspection and imaging techniques are used to acquire non-visible information embedded in samples. Typical applications include medical imaging, defect evaluation, and electronics testing. However, existing methods have specific limitations, including safety risks (e.g., X-ray), equipment costs (e.g., optical tomography), personnel training (e.g., ultrasonography) and material constraints (e.g., terahertz spectroscopy). Such constraints make these approaches impractical for everyday scenarios. In this paper, we present a method that is low-cost and practical for non-invasive inspection in everyday settings. Our prototype incorporates a miniaturized near-infrared spectroscopy scanner driven by a computer-controlled 2D-plotter. Our work presents a method to optimize content embedding, as well as a wavelength selection algorithm to extract content without human supervision. We show that our method can successfully extract occluded text through a paper stack of up to 16 pages. In addition, we present a deep-learning based image enhancement model that can further improve the image quality and simultaneously decompose overlapping content. Finally, we demonstrate how our method can be generalized to different inks and other layered materials beyond paper. Our approach enables a wide range of content embedding applications, including chipless information embedding, physical secret sharing, 3D print evaluations, and steganography.
Keywords: Gluten detection, NIRS, Prototyping, machine learning.
We investigate the use of a miniaturized Near-Infrared Spectroscopy (NIRS) device in a machine learning assisted decision-making task. We consider the real-world scenario of determining whether food contains gluten, and we investigate how end-users interact with our NIRS detection device to ultimately make this judgment. In particular, we explore the effects of different nutrition labels and representations of confidence on participants’ perception and trust. Our results show that participants tend to be conservative in their judgment and are willing to trust the device in the absence of understandable label information. We further identify strategies to increase user trust in the system. Our work contributes to the growing body of knowledge on how NIRS can be mass-appropriated for everyday sensing tasks, and how to enhance the trustworthiness of assisted decision-making systems.
Keywords: Liquid sensing, NIRS, Prototyping, machine learning.
Near-Infrared Spectroscopy (NIRS) is a non-invasive sensing technique which can be used to acquire information on an object's chemical composition. Although NIRS is conventionally used in dedicated laboratories, the recent introduction of miniaturized NIRS scanners has greatly expanded the use cases of this technology. Previous work shows that miniaturized NIRS can be successfully adapted to identify medical pills and alcohol concentration. We further extend this technology to identify sugar (sucrose) contents in everyday drinks. We developed a standalone mobile device which includes inter alia a NIRS scanner and a 3D printed clamp. The clamp can be attached to a straw-like tube to sense a liquid's sucrose content. Through a series of studies, we show that our technique can accurately measure sucrose levels in both lab-made samples and commercially available drinks, as well as classify commercial drinks. Furthermore, we show that our method is robust to variations in the ambient temperature and lighting conditions, as well as other liquids including milk, perfume and alcohols. Our method contributes to the development of everyday "food scanners" consumers.
Keywords: Wireless communication, Backscattering, Ultra low power, Prototyping.
We designed and implemented an ultra low power Wi-Fi backscattering development platform using the Texas Instrumental (TI) MSP432 solution. The development board includes GPIO ports for connecting data source devices such as sensors, microphones, and cameras. We also integrated an RF front-end using analog switches for data modulation and reflecting Wi-Fi signals. The size of the PCB board was similar to a credit card of 84 x 56 mm with 0.8 mm thickness, which can be further miniaturized using smaller parts. In addition, we also implemented a GNU Radio software for SDR hardware for receiver-side development.
Keywords: Wireless power transfer, Soft robot, Prototyping.
Wireless power transfer (WPT) has the significant potential for soft-bodied continuum robots to extend the operational time limitlessly and reduce weight. However, rigid power receiver coils, widely used in WPT, hinder the continuum deformation of the robot, and as a result, the function realization using the continuum deformation (e.g., locomotion) is impaired. Therefore, we introduce that a soft-bodied continuum robot can be designed by using thin film receiver coils and an inductively coupled wireless powering solution without sacrificing the continuum deformation and locomotion ability. A system is described for powering and controlling a soft robotic caterpillar consisting of nothing more than its continuum structure, actuators, and thin/flexible power receiving coils.
Keywords: Wireless power, Smart ring, Wearable.
Smart ring is one of the emerging wearables. It provides magnitudes of higher efficiency for input tasks, such as text input and gesture recognition. However, most smart rings suffer from large size due to the need of batteries, which deteriorates their wearability.
In this study, we present a truly wearable wirelessly powered smart rings that is thin, lightweight, and can work freely without battery life concerns. For achieving this functionality, our system aims at delivering power from a wristband to the smart rings wirelessly, utilizing the magnetic resonant coupling technique.
Keywords: Optimization, Information Theory, Wireless Communication, Algorithm, WSN.
In this study, we investigate the power allocation problem for an asymmetric wireless senor network, where multiple sensors observe a common binary source and transmit their corrupted observations to a data fusion node. We propose a power allocation scheme by maximizing the weighted channel capacity subject to the sum power constraint and show that this problem is convex. The simulation results verify that the proposed power allocation scheme outperforms the uniform power allocation method. Furthermore, a scheduling scheme for binary data gathering is proposed by determining the sensors that dominate the bit error rate performance.
Keywords: Wireless Communication, Prototyping, Smartphone.
We present Pulse, a wireless magnetic communication protocol for smartphones. Pulse is designed for off-the-shelf Android smartphones with magnetometers, and encodes data in magnetic fields. We present the design and evaluation of Pulse in various conditions (e.g., different voltages, number of transfer channels). The system provides security due to its short range (~1 cm), it can reach a speed of up to 44 bits per second, and it is possible to run it on most mobile phones with a magnetometer. We present our evaluation and discuss practical use cases where Pulse can be used today.
Keywords: UAV, Quadcopter, Control, Kalman filter, Web development (full-stack).
We present a collaboration system of ground based robot and UAVs forming a sensor network. The ground based robot can control multi UAVs with predefined scripts simultaneously. It can acquire environmental condition information (temperature, magnetism, and barometric pressure) and images from the UAVs and upload the data to a cloud server. A website is developed to provide interactive data visualization functions.
We also studied on Inertial Navigation System for the UAVs, the experimental results showed that the errors of Inertial Navigation System are very huge, thus future work for navigation and positioning is required by combining onboard vision.
Some random but fun projects and codings.
I somethings upload photos to Google Maps, here is the most popular one (more than 2 million views).