An official website of the United States government
The lack of adequate training data is one of the major hurdles in WiFi-based activity recognition systems. In this paper, we propose Wi-Fringe, which is a WiFi CSI-based devicefree human gesture recognition system that recognizes named gestures, i.e., activities and gestures that have a semantically meaningful name in English language, as opposed to arbitrary free-form gestures. Given a list of activities (only their names in English text), along with zero or more training examples (WiFi CSI values) per activity, Wi-Fringe is able to detect all activities at runtime. We show for the first time that by utilizing the state-of-the-art semantic representation of English words, which is learned from datasets like the Wikipedia (e.g., Google's word-to-vector [1]) and verb attributes learned from how a word is defined (e.g, American Heritage Dictionary), we can enhance the capability of WiFi-based named gesture recognition systems that lack adequate training examples per class. We propose a novel cross-domain knowledge transfer algorithm between radio frequency (RF) and text to lessen the burden on developers and end-users from the tedious task of data collection for all possible activities. To evaluate Wi-Fringe, we collect data from four volunteers in a multi-person apartment and an office building for a total of 20 activities. We empirically quantify the trade-off between the accuracy and the number of unseen activities.
more »
« less
This content will become publicly available on March 17, 2026
SandDune: Single ANtenna Device for Detecting User’s Natural Eating Habits
Over the years, researchers have explored various approaches for capturing and monitoring the eating activity, one among which is via Wi-Fi channel state information (CSI). CSI-based approaches commonly rely on multi-antenna systems for the capturing and monitoring tasks. With the advent of low-cost, single-antenna IoT devices with CSI measuring capabilities, a question that arises is whether these inexpensive devices can monitor human activities? In this paper we present the SandDune system that demonstrates the possibility of monitoring one human activity–eating–using only inexpensive single-antenna Wi-Fi devices. SandDune is an infrastructure-based system that continuously monitors CSI information to detect the eating activity occurring in its vicinity. When it detects an eating activity, it scrutinizes the signals further to identify all hand-to-mouth eating gestures in the eating episode. We tested SandDune and observed that SandDune can distinguish eating from other activities with an F1-score of 85.54%. Furthermore, it can detect the number of hand-to-mouth gestures that occurred in the eating episode with an error of ±3 gestures. Overall, we believe that a SandDune-like system can enable low cost, unobtrusive eating activity detection and monitoring with potential use-cases in several health and well-being applications.
more »
« less
- Award ID(s):
- 1955805
- PAR ID:
- 10618360
- Publisher / Repository:
- IEEE
- Date Published:
- ISBN:
- 979-8-3315-3553-7
- Page Range / eLocation ID:
- 689 to 692
- Format(s):
- Medium: X
- Location:
- Washington DC, DC, USA
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Vision-based methods are commonly used in robotic arm activity recognition. These approaches typically rely on line-of-sight (LoS) and raise privacy concerns, particularly in smart home applications. Passive Wi-Fi sensing represents a new paradigm for recognizing human and robotic arm activi- ties, utilizing channel state information (CSI) measurements to identify activities in indoor environments. In this paper, a novel machine learning approach based on discrete wavelet transform and vision transformers for robotic arm activity recognition from CSI measurements in indoor settings is proposed. This method outperforms convolutional neural network (CNN) and long short- term memory (LSTM) models in robotic arm activity recognition, particularly when LoS is obstructed by barriers, without relying on external or internal sensors or visual aids. Experiments are conducted using four different data collection scenarios and four different robotic arm activities. Performance results demonstrate that wavelet transform can significantly enhance the accuracy of visual transformer networks in robotic arms activity recognition.more » « less
-
null (Ed.)The development and validation of computational models to detect daily human behaviors (e.g., eating, smoking, brushing) using wearable devices requires labeled data collected from the natural field environment, with tight time synchronization of the micro-behaviors (e.g., start/end times of hand-to-mouth gestures during a smoking puff or an eating gesture) and the associated labels. Video data is increasingly being used for such label collection. Unfortunately, wearable devices and video cameras with independent (and drifting) clocks make tight time synchronization challenging. To address this issue, we present the Window Induced Shift Estimation method for Synchronization (SyncWISE) approach. We demonstrate the feasibility and effectiveness of our method by synchronizing the timestamps of a wearable camera and wearable accelerometer from 163 videos representing 45.2 hours of data from 21 participants enrolled in a real-world smoking cessation study. Our approach shows significant improvement over the state-of-the-art, even in the presence of high data loss, achieving 90% synchronization accuracy given a synchronization tolerance of 700 milliseconds. Our method also achieves state-of-the-art synchronization performance on the CMU-MMAC dataset.more » « less
-
null (Ed.)Detecting when eating occurs is an essential step toward automatic dietary monitoring, medication adherence assessment, and diet-related health interventions. Wearable technologies play a central role in designing unobtrusive diet monitoring solutions by leveraging machine learning algorithms that work on time-series sensor data to detect eating moments. While much research has been done on developing activity recognition and eating moment detection algorithms, the performance of the detection algorithms drops substantially when the model is utilized by a new user. To facilitate the development of personalized models, we propose PALS, Proximity-based Active Learning on Streaming data, a novel proximity-based model for recognizing eating gestures to significantly decrease the need for labeled data with new users. Our extensive analysis in both controlled and uncontrolled settings indicates F-score of PALS ranges from 22% to 39% for a budget that varies from 10 to 60 queries. Furthermore, compared to the state-of-the-art approaches, off-line PALS achieves up to 40% higher recall and 12% higher F-score in detecting eating gestures.more » « less
-
Circumvent traffic shaping using virtual wireless clients in IEEE 802.11 wireless local area networkAccessing the Internet through Wi-Fi networks offers an inexpensive alternative for offloading data from mobile broadband connections. Businesses such as fast food restaurants, coffee shops, hotels, and airports, provide complimentary Internet access to their customers through Wi-Fi networks. Clients can connect to the Wi-Fi hotspot using different wireless devices. However, network administrators may apply traffic shaping to control the wireless client's upload and download data rates. Such limitation is used to avoid overloading the hotspot, thus providing fair bandwidth allocation. Also, it allows for the collection of money from the client in order to have access to a faster Internet service. In this paper, we present a new technique to avoid bandwidth limitation imposed by Wi-Fi hotspots. The proposed method creates multiple virtual wireless clients using only one physical wireless interface card. Each virtual wireless client emulates a standalone wireless device. The combination of the individual bandwidth of each virtual wireless client results in an increase of the total bandwidth gained by the attacker. Our proposed technique was implemented and evaluated in a real-life environment with an increase in data rate up to 16 folds.more » « less
