Kernel fusion based extreme learning machine for cross-location activity recognition

Wang, Zhelong; Wu, Donghui; Gravina, Raffaele; Fortino, Giancarlo; Jiang, Yongmei; Tang, Kai

doi:10.1016/j.inffus.2017.01.004

Cited by 82 publications

(26 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wang et al formulated an approach based on extreme learning machines [32]. They focused on a set of four IMUs that were distributed on a subject's body's right ankle, waist, wrist and right on top of the bellybutton.…”

Section: Related Workmentioning

confidence: 99%

“…It can also be concluded that most approaches have an idea on the expansion of a system to cover further input sources, however, they either rely on ground truth labels (something that is not available in real life scenarios; cf. [32]) make assumptions that do not hold up on real life datasets or describe systems with narrowly defined rules for integrating/replacing sensors in a very restrictive manner [30]. Another gap we want to overcome is the missing link between observed structure in data (captured via generative models) and the addressed classification task (discriminative models).…”

Section: Research Gapmentioning

confidence: 99%

See 1 more Smart Citation

Self-Adaptive Multi-Sensor Activity Recognition Systems Based on Gaussian Mixture Models

2018

View full text Add to dashboard Cite

Personal wearables such as smartphones or smartwatches are increasingly utilized in everyday life. Frequently, activity recognition is performed on these devices to estimate the current user status and trigger automated actions according to the user’s needs. In this article, we focus on the creation of a self-adaptive activity recognition system based on IMU that includes new sensors during runtime. Starting with a classifier based on GMM, the density model is adapted to new sensor data fully autonomously by issuing the marginalization property of normal distributions. To create a classifier from that, label inference is done, either based on the initial classifier or based on the training data. For evaluation, we used more than 10 h of annotated activity data from the publicly available PAMAP2 benchmark dataset. Using the data, we showed the feasibility of our approach and performed 9720 experiments, to get resilient numbers. One approach performed reasonably well, leading to a system improvement on average, with an increase in the F-score of 0.0053, while the other one shows clear drawbacks due to a high loss of information during label inference. Furthermore, a comparison with state of the art techniques shows the necessity for further experiments in this area.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Research Gapmentioning

confidence: 99%

Self-Adaptive Multi-Sensor Activity Recognition Systems Based on Gaussian Mixture Models

2018

View full text Add to dashboard Cite

show abstract

“…Hence, a significant amount of research is still required to develop algorithms that can help with an accurate diagnosis [42,43,44]. Moreover, frequent patterns are huge in space if we try to find them in BSN-generated data and are limited in terms of detecting changes in human behavior that occur regularly or periodic in everyday life.…”

Section: Related Workmentioning

confidence: 99%

Mining Productive-Associated Periodic-Frequent Patterns in Body Sensor Data for Smart Home Care

Ismail

Hassan

2017

Sensors

View full text Add to dashboard Cite

The understanding of various health-oriented vital sign data generated from body sensor networks (BSNs) and discovery of the associations between the generated parameters is an important task that may assist and promote important decision making in healthcare. For example, in a smart home scenario where occupants’ health status is continuously monitored remotely, it is essential to provide the required assistance when an unusual or critical situation is detected in their vital sign data. In this paper, we present an efficient approach for mining the periodic patterns obtained from BSN data. In addition, we employ a correlation test on the generated patterns and introduce productive-associated periodic-frequent patterns as the set of correlated periodic-frequent items. The combination of these measures has the advantage of empowering healthcare providers and patients to raise the quality of diagnosis as well as improve treatment and smart care, especially for elderly people in smart homes. We develop an efficient algorithm named PPFP-growth (Productive Periodic-Frequent Pattern-growth) to discover all productive-associated periodic frequent patterns using these measures. PPFP-growth is efficient and the productiveness measure removes uncorrelated periodic items. An experimental evaluation on synthetic and real datasets shows the efficiency of the proposed PPFP-growth algorithm, which can filter a huge number of periodic patterns to reveal only the correlated ones.

show abstract

“…Real-time human motion recognition is useful for fall detection in home care research [13]. Machine learning techniques can be applied to longitudinal recording of human motion patterns to allow for more precise classification and recognition of various movement types [14]. Some studies have focused on the evaluation of hemiplegia patients [15,16] using machine learning techniques to train samples for the development of classification and evaluation models.…”

Section: Introductionmentioning

confidence: 99%

A Brunnstrom Stage Evaluation System with Fuzzy-Accelerometer

Tsai

et al. 2019

Applied Sciences

View full text Add to dashboard Cite

Advances in medical care has reduced the rate of mortality from strokes, but the incidence of stroke has remained stable while the incidence of ministrokes has increased. Most stroke victims require long-term care, imposing a heavy financial and emotional burden on families while incurring a heavy cost to society. Thus, strokes are a key issue in the context of health care in Taiwan. This paper proposes using VBA (Visual Basic for Applications) to build a system for assessing Brunnstrom stages based on the observation of several obvious rehabilitation features The system calculates features for accelerometer readings, which are then used as input parameters for a fuzzy algorithm to obtain the Brunnstrom action level. Experimental results show the proposed approach effectively assesses Brunnstrom level, and that the approach can be used to assist physical therapists in performing longitudinal assessments of stroke victim progress, thus improving evaluation efficiency.

show abstract

Kernel fusion based extreme learning machine for cross-location activity recognition

Cited by 82 publications

References 31 publications

Self-Adaptive Multi-Sensor Activity Recognition Systems Based on Gaussian Mixture Models

Self-Adaptive Multi-Sensor Activity Recognition Systems Based on Gaussian Mixture Models

Mining Productive-Associated Periodic-Frequent Patterns in Body Sensor Data for Smart Home Care

A Brunnstrom Stage Evaluation System with Fuzzy-Accelerometer

Contact Info

Product

Resources

About