Stratified Transfer Learning for Cross-domain Activity Recognition

Wang, Jindong; Chen, Yiqiang; Hu, Lisha; Peng, Xiaohui; Yu, Philip S.

doi:10.1109/percom.2018.8444572

Cited by 183 publications

(116 citation statements)

References 39 publications

Supporting

Mentioning

106

Contrasting

Order By: Relevance

“…• STL: Strati ed Transfer Learning [40]. PCA and KPCA are classic dimensionality reduction methods, while TCA, GFK, TKL, and STL are representative transfer learning approaches.…”

Section: Evaluation Of Tnnarmentioning

confidence: 99%

“…e constructed datasets and STL code are available online 2 . e implementations of all comparison methods are following [40]. Di erent from these work which exploited feature extraction according to human knowledge, we take the original signal as the input.…”

Section: Evaluation Of Tnnarmentioning

confidence: 99%

“…Note that in both of the two experiments, all of the comparison methods used the same source and target domains. For the single source domain situation, we follow the se ings in [40] and report the results in Table 2. For the experiments on multiple source domains, we extend the results in the last section and set K = 3 to select the source domains by USSAR.…”

Section: Evaluation Of Tnnarmentioning

confidence: 99%

See 2 more Smart Citations

Deep Transfer Learning for Cross-domain Activity Recognition

Wang

Zheng

Chen

et al. 2018

Proceedings of the 3rd International Conference on Crowd Science and Engineering

Self Cite

115

View full text Add to dashboard Cite

Human activity recognition plays an important role in people's daily life. However, it is o en expensive and time-consuming to acquire su cient labeled activity data. To solve this problem, transfer learning leverages the labeled samples from the source domain to annotate the target domain which has few or none labels. Unfortunately, when there are several source domains available, it is di cult to select the right source domains for transfer. e right source domain means that it has the most similar properties with the target domain, thus their similarity is higher, which can facilitate transfer learning. Choosing the right source domain helps the algorithm perform well and prevents the negative transfer. In this paper, we propose an e ective Unsupervised Source Selection algorithm for Activity Recognition (USSAR). USSAR is able to select the most similar K source domains from a list of available domains. After this, we propose an e ective Transfer Neural Network to perform knowledge transfer for Activity Recognition (TNNAR). TNNAR could capture both the time and spatial relationship between activities while transferring knowledge. Experiments on three public activity recognition datasets demonstrate that: 1) e USSAR algorithm is e ective in selecting the best source domains. 2) e TNNAR method can reach high accuracy when performing activity knowledge transfer.

show abstract

“…• STL: Strati ed Transfer Learning [40]. PCA and KPCA are classic dimensionality reduction methods, while TCA, GFK, TKL, and STL are representative transfer learning approaches.…”

Section: Evaluation Of Tnnarmentioning

confidence: 99%

Section: Evaluation Of Tnnarmentioning

confidence: 99%

Section: Evaluation Of Tnnarmentioning

confidence: 99%

See 1 more Smart Citation

Deep Transfer Learning for Cross-domain Activity Recognition

Wang

Zheng

Chen

et al. 2018

Proceedings of the 3rd International Conference on Crowd Science and Engineering

Self Cite

115

View full text Add to dashboard Cite

show abstract

“…Transfer learning (TL), or domain adaptation [12] is a promising strategy to enhance the learning performance on a target domain with few or none labels by leveraging knowledge from a well-labeled source domain. Since the source and target domains have dif- ferent distributions, numerous methods have been proposed to reduce the distribution divergence [24,23,21,26,6]. Unfortunately, despite the great success achieved by existing TL methods, it is notoriously challenging to apply them to a real situation since we cannot determine the best TL model and their optimal hyperparameters.…”

Section: Introductionmentioning

confidence: 99%

Easy Transfer Learning By Exploiting Intra-Domain Structures

Wang

Chen

et al. 2019

2019 IEEE International Conference on Multimedia and Expo (ICME)

Self Cite

126

View full text Add to dashboard Cite

Transfer learning aims at transferring knowledge from a welllabeled domain to a similar but different domain with limited or no labels. Unfortunately, existing learning-based methods often involve intensive model selection and hyperparameter tuning to obtain good results. Moreover, cross-validation is not possible for tuning hyperparameters since there are often no labels in the target domain. This would restrict wide applicability of transfer learning especially in computationallyconstraint devices such as wearables. In this paper, we propose a practically Easy Transfer Learning (EasyTL) approach which requires no model selection and hyperparameter tuning, while achieving competitive performance. By exploiting intra-domain structures, EasyTL is able to learn both nonparametric transfer features and classifiers. Extensive experiments demonstrate that, compared to state-of-the-art traditional and deep methods, EasyTL satisfies the Occam's Razor principle: it is extremely easy to implement and use while achieving comparable or better performance in classification accuracy and much better computational efficiency. Additionally, it is shown that EasyTL can increase the performance of existing transfer feature learning methods.

show abstract

“…This paper is an extended version of our PerCom paper [20], where we proposed a stratified transfer learning algorithm for activity transfer. That algorithm is regarded as STL-SAT in this paper.…”

Section: Introductionmentioning

confidence: 99%

Cross-position activity recognition with stratified transfer learning

Chen,

Wang,

Huang

et al. 2019

Pervasive and Mobile Computing

Self Cite

View full text Add to dashboard Cite

Human activity recognition (HAR) aims to recognize the activities of daily living by utilizing the sensors attached to different body parts. The great success of HAR has been achieved by training machine learning models using sufficient labeled activity data. However, when the labeled data from a certain body position (i.e. target domain) is missing, how to leverage the data from other positions (i.e. source domain) to help recognize the activities of this position? This problem can be divided into two steps. Firstly, when there are several source domains available, it is often difficult to select the most similar source domain to the target domain. Secondly, with the selected source domain, we need to perform accurate knowledge transfer between domains in order to recognize the activities on the target domain. Existing methods only learn the global distance between domains while ignoring the local property. In this paper, we propose a Stratified Transfer Learning (STL) framework to perform both source domain selection and activity transfer. STL is based on our proposed Stratified distance to capture the local property of domains. STL consists of two components: 1) Stratified Domain Selection (STL-SDS), which can select the most similar source domain to the target domain; and 2) Stratified Activity Transfer (STL-SAT), which is able to perform accurate knowledge transfer. Extensive experiments on three public activity recognition datasets demonstrate the superiority of STL. Furthermore, we extensively investigate the performance of transfer learning across different degrees of similarities and activity levels between domains. We also discuss the potential applications of STL in other fields of pervasive computing for future research.

show abstract

Stratified Transfer Learning for Cross-domain Activity Recognition

Cited by 183 publications

References 39 publications

Deep Transfer Learning for Cross-domain Activity Recognition

Deep Transfer Learning for Cross-domain Activity Recognition

Easy Transfer Learning By Exploiting Intra-Domain Structures

Cross-position activity recognition with stratified transfer learning

Contact Info

Product

Resources

About