Collaborative Human Activity Recognition with Passive Inter-Body Electrostatic Field

Sizhen Bian, Vitor Fortes Rey, Siyu Yuan, Paul Lukowicz · 2025

Read Collaborative Human Activity Recognition with Passive Inter-Body Electrostatic Field by Sizhen Bian, Vitor Fortes Rey, Siyu Yuan, Paul Lukowicz (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-06940.txt.

Introduction

Sizhen Bian and three co-authors sense shared activities through the body's electric field. Their 2025 paper tests inter-body electrostatic sensing with wearables. The body's passive field reacts to full-body motion and to surroundings at once. Inertial sensors, by contrast, miss the surroundings entirely.

Wrist prototypes captured sixteen hours of labeled object handling. People worked alone and together on the tasks. A regression model sorted single and joint activities from the field. Against accelerometers, the new signal underperformed overall, to the team's surprise. Yet merging both signals lifted collaboration detection by sixteen percent. The field's gift is social, sensing when bodies work as one.

Catalog record

CREATORSizhen Bian, Vitor Fortes Rey, Siyu Yuan, Paul Lukowicz
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE24 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-06940

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Collaborative Human Activity Recognition with Passive Inter-Body Electrostatic Field Sizhen Bian, Vitor Fortes Rey, Siyu Yuan, Paul Lukowicz Abstract The passive body-area electrostatic field has recently been aspiringly explored for wearable motion sensing, harnessing its two thrilling characteristics: full-body motion sensitivity and environmental sensitivity, which potentially empowers human activity recognition both independently and jointly from a single sensing front-end and theoretically brings significant competition against traditional inertial sensor that is incapable in environmental variations sensing. While most works focus on exploring the electrostatic field of a single body as the target, this work, for the first time, quantitatively evaluates the mutual effect of inter-body electrostatic fields and its contribution to collaborative activity recognition. A wearable electrostatic field sensing front-end and wrist-worn prototypes are built, and a sixteen-hour, manually annotated dataset is collected, involving an experiment of manipulating objects both independently and collaboratively. A regression model is finally used to recognize the collaborative activities among users. Despite the theoretical advantages of the body electrostatic field, the recognition of both single and collaborative activities shows unanticipated less-competitive recognition performance compared with the accelerometer. However, It is worth mentioning that this novel sensing modality improves the recognition F-score of user collaboration by 16% in the fusion result of the two wearable motion sensing modalities, demonstrating the potential of bringing body electrostatic field as a complementary power-efficient signal for collaborative activity tracking using wearables. I Introduction Wearable sensors are increasingly captivating researchers for a better understanding of human behaviours and correspondingly provide assistive services [ 1 , 2 , 3 ] . Recent advances in wearable sensors have enabled a broader range of functional support in the field like sport [ 4 ] , healthcare [ 5 , 6 ] , clinic [ 7 ] , gaming [ 8 ] , interaction [ 9 , 10 ] , etc. Among them, wearable motion sensing has long been dominated by inertial sensors like accelerometers and gyroscopes, which play almost the unique role for motion sensing in current wearable devices [ 11 , 12 ] . However, one limitation of an inertial sensor is that it primarily perceives the motion of the body part to which it is attached and is limited in surrounding variation sensing (incapable) and cross-body part sensing (weak). In cross-body part sensing, the motion message of the targeted body part gradually gets lost when reaching the sensor-attached body part, e.g., the inaccuracy of step counting by a wrist-worn inertial sensor [ 13 , 14 ] . To this end, a novel wearable motion sensing technique has emerged to supply complementary sensing ability to the inertial sensor, the body-area electrostatic field, also named human body capacitance, inferring body motions by observing the body surface charge flow pattern, which strongly relates to the body motion pattern [ 15 ] . Unlike other physiological features like electrocardiograph, the body-area electrostatic field is a feature that interacts with the surroundings, especially the ground. In substance, the body-area electrostatic field describes the charge distribution difference between the body surface and the surroundings. When inferring motion information from this feature, the field shows two thrilling characteristics: full-body movement detection, meaning the deployment of the sensing unit on the targeted body part is not mandatory. For example, a wrist-worn body electrostatic field sensing unit can track leg-based exercises even when the wrist is absolutely static [ 16 ] ; and surrounding motion detection, where the field sensing unit is sensitive to the body’s interaction with its surroundings, including both contacting and being in the immediate proximity of people and objects [ 17 ] . Despite its subtlety, researchers have tried to make use of this intrinsic property for potential well-being services. Such explorations go back to early work on capacitive intra-body personal area networks [ 18 ] , followed by proximity sensing [ 19 ] , communication [ 20 ] , human-machine interaction [ 21 ] and motion detection [ 15 , 22 , 23 ] . Nevertheless, most existing works exploring wearable body electrostatic field sensing only focus on single-body activity perception or body-machine interaction, and body-to-body cooperation was not attentively explored. When multiple bodies collaboratively take action and are in the near range of each other, their body-area electrostatic field will be mutually affected; thus, extra collaborative contextual knowledge could be potentially deduced besides the traditional wearable motion sensing, leveraging the feature of full-body and surrounding sensitivity of the body electrostatic field. Based on this expectation, we moved one step further from the existing works of recognition of physical actions performed by an individual to physical actions collaboratively performed by group users [ 24 ] and tried to recognize physical activities that are a challenge for the classical wearable motion sensor. II Related Works Existing works in collaborative activity recognition were mainly based on camera surveillance videos, which focused on spatiotemporal relations among the human bodies in the scene and deduced collaborative activities through tracking multi-agent spots [ 25 , 26 ] . Video-based group activity recognition suffers from its high computational cost and other ethical issues. This work focuses on wearable sensor-based solutions and aims to facilitate the detection of joint physical manipulation of objects that are more difficult to detect visually. Concerning the sensor-based analysis of multi-agents activities, Wilson et al. [ 27 ] offered a combination of motion detectors, break-beam sensors, pressure mats, and contact switches to detect agents’ proximity and touch actions. An infrared sensor-based scalable network [ 28 ] was also proposed for perceiving people’s activities in an intelligent building and got better than 90% recognition performance for low-level activities. Wang et al. [ 29 ] developed a multi-modal wearable sensing platform and presented a theoretical framework to recognize both single-user and multi-user activities, in which temperature, humidity, light, audio, RFID, motion sensors were collected to contribute to the recognition. Gordon et al. [ 30 ] explored a node with motion sensors, which can be attached to a mug, by deploying several nodes on mugs; the collaborative group activity was extracted at mobile devices’ side. Those multi-modal sensor-based multi-agent activity recognition platforms were heavy to deploy, and they usually focused on detecting whether the involved agents were gathering or dispersing as an indication of group activities. A high-level activity like agent collaboration is a more challenging topic and exists in plenty of practical scenarios, especially in manufacturing [ 31 , 32 , 33 ] . Fig. 1 : Example of the type of recognition tasks that our work partly targets: distinguishing group agents among normal walking next to each other, walking next to each other and operating an object alone, and walking next to each other while jointly operating an object. Co-location is not a solid separator, and camera filming is often limited by line of sight. Benefiting from the intrinsic property of the body-area electrostatic field and its low power and small form factor sensing unit, this work explores this novel sensing modality for recognizing the typical scenario of physical works related to manipulating objects and physical collaboration between users. The problem we addressed is partly illustrated in Figure 1 . Previous work has investigated using inertial sensors and co-location information for such recognition tasks [ 34 ] . The problem

Read the complete text