Search for: All records

Visual body signals are designated body poses that deliver an application-specific message. Such signals are widely used for fast message communication in sports (signaling by umpires and referees), transportation (naval officers and aircraft marshallers), and construction (signaling by riggers and crane operators), to list a few examples. Automatic interpretation of such signals can help maintaining safer operations in these industries, help in record-keeping for auditing or accident investigation purposes, and function as a score-keeper in sports. When automation of these signals is desired, it is traditionally performed from a viewer's perspective by running computer vision algorithms on camera feeds. However, computer vision based approaches suffer from performance deterioration in scenarios such as lighting variations, occlusions, etc., might face resolution limitations, and can be challenging to install. Our work, ViSig, breaks with tradition by instead deploying on-body sensors for signal interpretation. Our key innovation is the fusion of ultra-wideband (UWB) sensors for capturing on-body distance measurements, inertial sensors (IMU) for capturing orientation of a few body segments, and photodiodes for finger signal recognition, enabling a robust interpretation of signals. By deploying only a small number of sensors, we show that body signals can be interpreted unambiguously in many different settings, including in games of Cricket, Baseball, and Football, and in operational safety use-cases such as crane operations and flag semaphores for maritime navigation, with > 90% accuracy. Overall, we have seen substantial promise in this approach and expect a large body of future follow-on work to start using UWB and IMU fused modalities for the more general human pose estimation problems. 

High-precision tracking of a pen-like instrument's movements is desirable in a wide range of fields spanning education, robotics, and art, to name a few. The key challenge in doing so stems from the impracticality of embedding electronics in the tip of such instruments (a pen, marker, scalpel, etc.) as well as the difficulties in instrumenting the surface that it works on. In this paper, we present ITrackU, a movement digitization system that does not require modifications to the surface or the tracked instrument's tip. ITrackU fuses locations obtained using ultra-wideband radios (UWB), with an inertial and magnetic unit (IMU) and a pressure sensor, yielding multidimensional improvements in accuracy, range, cost, and robustness, over existing works. ITrackU embeds a micro-transmitter at the base of a pen which creates a trackable beacon, that is localized from the corners of a writing surface. Fused with inertial motion sensor and a pressure sensor, ITrackU enables accurate tracking. Our prototype of ITrackU covers a large 2.5m × 2m area, while obtaining around 2.9mm median error. We demonstrate the accuracy of our system by drawing numerous shapes and characters on a whiteboard, and compare them against a touchscreen and a camera-based ground-truthing system. Finally, the produced stream of digitized data is minuscule in volume, when compared with a video of the whiteboard, which saves both network bandwidth and storage space. 

In this poster, we present the potential of extending 6Fit-a-Part, our recently proposed physical distancing platform, to enable interactive physical games in school campuses despite COVID-19 restrictions. To minimize the risk of infection, traditional physical games must be modified such that the inter-player distance remains beyond 6 feet at all times. Our wearable electronic gadget that beeps when it approaches another similar device can facilitate such games, however, it must first solve 3 fundamental challenges: high accuracy, low delay, and high robustness. We highlight that 6Fit-a-Part adopts an improved two way ranging protocol using ultra-wideband radio (UWB) which can provide accurate inter-player distance measurements in real-time. Furthermore, 6Fit-a-Part leverages wireless channel features to perform occlusion detection continuously so that erroneous measurements caused by human occlusions can be corrected. 6Fit-a-Part is designed to be a lightweight wearable device making it a suitable accessory even during games. By compromising rules of tradition physical games, we show that 6Fit-a-Part is capable of seamlessly re-enabling physical games while still enabling physical distancing. 

The coronavirus pandemic is altering our way of life. As more establishments open, there is an expectation that people will follow physical distancing guidelines. The implementation, however, is poor; just putting up warning signs appealing the general public to keep a distance of 6 feet from others is hardly enough. In this paper we consider the design of a wearable device that raises an alarm if another similar device is detected within a set distance. It uses off-the-shelf ultra-wideband radio technology for real-time, accurate distance estimation from others in the vicinity. We design an one-to-all ranging protocol that is able to accurately estimate distance to neighboring devices and warn the user if the distance falls below a certain established threshold within a short time. The device must compensate for human occlusions and avoid unnecessary warnings when physical barriers exist between devices. We implement and evaluate our protocol in a small testbed with custom prototype hardware as well as in simulation. Our ranging protocol is capable of performing up to 10 distance measurements per second, while avoiding packet collisions. The overall percentage of rangings completed is around 65% in a 10-node network, and the distance accuracy is around 20cm even with frequent human occlusions. We believe this prototype will provide the first steps to ensure physical distancing in various real-world settings.