2016 IEEE International Conference on Pervasive Computing and Communications (PerCom) 2016
DOI: 10.1109/percom.2016.7456502
|View full text |Cite
|
Sign up to set email alerts
|

Adaptive activity learning with dynamically available context

Abstract: Numerous methods have been proposed to address different aspects of human activity recognition. However, most of the previous approaches are static in terms of the data sources used for the recognition task. As sensors can be added or can fail and be replaced by different types of sensors, creating an activity recognition model that is able to leverage dynamically available sensors becomes important. In this paper, we propose methods for activity learning and activity recognition adaptation in environments wit… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
14
0

Year Published

2018
2018
2023
2023

Publication Types

Select...
5
1
1

Relationship

1
6

Authors

Journals

citations
Cited by 20 publications
(14 citation statements)
references
References 28 publications
0
14
0
Order By: Relevance
“…Our approach to sensor-context learning is a hybrid method that combines MoE and transfer learning in a unified framework. Although prior studies combine sensor localization or context detection within activity classification [24,31], they require a set of labeled data to train the localization/context-learning model. To the best of our knowledge, there is no prior research that addresses the problem of the MoE model considering no available ground truth and only relying on the performance of available experts for wearable sensors.…”
Section: Comparative Evaluation Methodsmentioning
confidence: 99%
See 1 more Smart Citation
“…Our approach to sensor-context learning is a hybrid method that combines MoE and transfer learning in a unified framework. Although prior studies combine sensor localization or context detection within activity classification [24,31], they require a set of labeled data to train the localization/context-learning model. To the best of our knowledge, there is no prior research that addresses the problem of the MoE model considering no available ground truth and only relying on the performance of available experts for wearable sensors.…”
Section: Comparative Evaluation Methodsmentioning
confidence: 99%
“…As soon as the location of the sensor is identified, one can use an activity recognition model specifically trained for the detected wearing site to infer human activities. Wen et al [31] proposed a probabilistic context modeling that uses ambient sensors to detect sensor context. Then, they used the context model to rank possible activity labels.…”
Section: Sensor Localizationmentioning
confidence: 99%
“…Human activity recognition (HAR) is a hot research topic in pervasive computing. HAR has been widely applied to many applications such as indoor localization [1], sleep state detection [2], and smart home sensing [3]. The key to successful HAR is to build accurate and robust models using sufficient labeled activity data.…”
Section: Introductionmentioning
confidence: 99%
“…For example, we can detect if a person is walking or running using the on-body sensors such as the smartphone or the wristband. HAR has been widely used to applications such as smart care [2], wireless sensing [3,4], adaptive systems [5], and smart home sensing [6,7].…”
Section: Introductionmentioning
confidence: 99%