Combining Motion Sensors and Ultrasonic Hands Tracking for Continuous Activity Recognition in a Maintenance Scenario

Stiefmeier, Thomas; Ogris, Georg; Junker, Holger; Lukowicz, Paul; Tröster, Gerhard

doi:10.1109/iswc.2006.286350

Cited by 69 publications

(37 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Watanabe et al [26] developed an activity and context recognition method where the user carries a neck-worn receiver comprising a microphone and small speakers on his/her wrists that generate ultrasounds. Stiefmeier et al [24] presented a method for continuous activity recognition based on ultrasonic hand tracking and motion sensors attached to the user's arms. Gupta et al [13] presented SoundWave, a technique that leverages the speaker and microphone already embedded in most commodity devices to sense in-air gestures around the device.…”

Section: Gesture Recognition Methodsmentioning

confidence: 99%

Gesture Recognition Method Utilizing Ultrasonic Active Acoustic Sensing

Watanabe

Terada

Tsukamoto

2017

Journal of Information Processing

View full text Add to dashboard Cite

Abstract:We propose a method for gesture recognition that utilizes active acoustic sensing, which transmits acoustic signals to a target, and recognizes the target's state by analyzing the response. In this study, the user wore a contact speaker that transmitted ultrasonic sweep signals to the user's body and a contact microphone that detected the ultrasound propagated through the body. The propagation characteristics of the ultrasound changed depending on the user's movements. We utilized these changes to recognize the user's gestures. One of the important novelty features of our method is that the user's gestures can be acquired not only from the physical movement but also from the user's internal state, such as muscle activity, since ultrasound is transmitted via both the user's internal body and body surface. Moreover, our method is not adversely affected by audible-range sounds generated by the environment and body movements because we utilize ultrasound. We implemented a device that uses active acoustic sensing to effectively transmit/detect the ultrasound to/from the body and investigated the performance of the proposed method in 21 contexts with 10 subjects. The evaluation results confirmed that the precision and recall are 93.1% and 91.6%, respectively when we set 10% of the data as training data and the rest as testing data in the same data set. When we used the data set for training and the other data set for testing in the same day, the precision and recall are 51.6% and 51.3%, respectively.

show abstract

Section: Gesture Recognition Methodsmentioning

confidence: 99%

Gesture Recognition Method Utilizing Ultrasonic Active Acoustic Sensing

Watanabe

Terada

Tsukamoto

2017

Journal of Information Processing

View full text Add to dashboard Cite

show abstract

“…The dataset contains 20 activities that are performed during a typical car quality inspection [22]. Example activities are checking gaps of the car's body or inspecting movable parts, for instance, by opening and closing doors.…”

Section: B Car Quality Inspectionmentioning

confidence: 99%

All for one or one for all? Combining heterogeneous features for activity spotting

Blanke

Schiele

Kreil

et al. 2010

2010 8th IEEE International Conference on Pervasive Computing and Communications Workshops (PERCOM Workshops)

Self Cite

View full text Add to dashboard Cite

Abstract-Choosing the right feature for motion based activity spotting is not a trivial task. Often, features derived by intuition or that proved to work well in previous work are used. While feature selection algorithms allow automatic decision, definition of features remains a manual task. We conduct a comparative study of features with very different origin. To this end, we propose a new type of features based on polynomial approximation of signals. The new feature type is compared to features used routinely for motion based activity recognition as well as to recently proposed body-model based features. Experiments were performed on three different, large datasets allowing a thorough, in-depth analysis. They not only show the respective strengths of the different feature types but also their complementarity resulting in improved performance through combination. It shows that each feature type with its individual and complementary strengths and weaknesses can improve results by combination.

show abstract

“…A variety of methods and algorithms have been developed to learn valuable information 5 from the data, and to make the manufacturing smarter [2]. The fast-growing artificial intelligence technologies, particularly deep learning [3], are promising to further boost this industry.…”

Section: Introductionmentioning

confidence: 99%

“…The fast-growing artificial intelligence technologies, particularly deep learning [3], are promising to further boost this industry. In a smart manufacturing system involving workers, recognition of the worker's activity can be used for quantification and evaluation of the worker's performance, as well as to provide onsite instructions with For activity recognition in the manufacturing area, Stiefmeire et al [5] utilized ultrasonic and IMU sensors for worker activity recognition in a bicycle maintenance scenario using a Hidden Markov Model classifier. Later they proposed a string-matching based segmentation and classification method us-20 ing multiple IMU sensors for recognizing worker activity in car manufacturing tasks [6,7].…”

Section: Introductionmentioning

confidence: 99%

Worker Activity Recognition in Smart Manufacturing Using IMU and sEMG Signals with Convolutional Neural Networks

Tao

Lai

Leu

et al. 2018

EasyChair Preprints

View full text Add to dashboard Cite

In a smart manufacturing system involving workers, recognition of the worker's activity can be used for quantification and evaluation of the worker's performance, as well as to provide onsite instructions with augmented reality. In this paper, we propose a method for activity recognition using Inertial Mea- and achieves 98% and 87% recognition accuracy in the half-half and leave-oneout experiments, respectively.

show abstract

Combining Motion Sensors and Ultrasonic Hands Tracking for Continuous Activity Recognition in a Maintenance Scenario

Cited by 69 publications

References 19 publications

Gesture Recognition Method Utilizing Ultrasonic Active Acoustic Sensing

Gesture Recognition Method Utilizing Ultrasonic Active Acoustic Sensing

All for one or one for all? Combining heterogeneous features for activity spotting

Worker Activity Recognition in Smart Manufacturing Using IMU and sEMG Signals with Convolutional Neural Networks

Contact Info

Product

Resources

About