Cross-domain activity recognition via transfer learning

Hu, Derek Hao; Zheng, Vincent W.; Yang, Qiang

doi:10.1016/j.pmcj.2010.11.005

Cited by 77 publications

(45 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1 A snapshot from the event the setup of the transfer differs from our work. In their survey, Cook et al [6] grouped existing transfer learning studies with respect to the modalities used: video sequences [34], wearable [11,37] and ambient sensors [16].…”

Section: Transfer Learning For Behaviour Recognitionmentioning

confidence: 99%

See 1 more Smart Citation

Personalised models for speech detection from body movements using transductive parameter transfer

Gedik

Hung

2017

Pers Ubiquit Comput

View full text Add to dashboard Cite

We investigate the task of detecting speakers in crowded environments using a single body worn triaxial accelerometer. Detection of such behaviour is very challenging to model as people's body movements during speech vary greatly. Similar to previous studies, by assuming that body movements are indicative of speech, we show experimentally, on a real-world dataset of 3 h including 18 people, that transductive parameter transfer learning (Zen et al. in Proceedings of the 16th international conference on multimodal interaction. ACM, 2014) can better model individual differences in speaking behaviour, significantly improving on the state-of-the-art performance. We also discuss the challenges introduced by the in-thewild nature of our dataset and experimentally show how they affect detection performance. We strengthen the need for an adaptive approach by comparing the speech detection problem to a more traditional activity (i.e. walking). We provide an analysis of the transfer by considering different source sets which provides a deeper investigation of the nature of both speech and body movements, in the context of transfer learning.

show abstract

Section: Transfer Learning For Behaviour Recognitionmentioning

confidence: 99%

“…For example, Hu et al [11] proposed a method, which focused on cross-domain activity recognition. They transferred the information from an available labelled data of a set of existing activities to a different yet still related set of activities.…”

Section: Transfer Learning For Behaviour Recognitionmentioning

confidence: 99%

Personalised models for speech detection from body movements using transductive parameter transfer

Gedik

Hung

2017

Pers Ubiquit Comput

View full text Add to dashboard Cite

show abstract

“…Authors in [41] create a codebook with unlabelled data from the source domain and train the recognition with labelled data from the target domain, being an informed unsupervised (IU) transfer learning method. The literature is really extensive, but most of the methods use IEEE TRANSACTIONS ON CYBERNETICS 4 similar approaches [42] [43]. The UBM vocabulary adaptation proposed in this paper lays in the US transfer learning as the source domain descriptors are labelled in a GMM training process and this GMM is later adapted to unlabelled target domain descriptors.…”

Section: Related Workmentioning

confidence: 99%

One-Shot Learning of Human Activity With an MAP Adapted GMM and Simplex-HMM

Rodriguez

Orrite

Medrano

et al. 2017

IEEE Trans. Cybern.

View full text Add to dashboard Cite

Abstract-This paper presents a novel activity class representation using a single sequence for training. The contribution of this representation lays on the ability to train an one-shot learning recognition system, useful in new scenarios where capturing and labelling sequences is expensive or impractical. The method uses a Universal Background Model of local descriptors obtained from source databases available on-line and adapts it to a new sequence in the target scenario through a Maximum a Posteriori adaptation. Each activity sample is encoded in a sequence of normalized Bag of Features and modelled by a new Hidden Markov Model formulation, where the EM algorithm for training is modified to deal with observations consisting in vectors in a unit simplex. Extensive experiments in recognition have been performed using one-shot learning over the public datasets Weizmann, KTH and IXMAS. These experiments demonstrate the discriminative properties of the representation and the validity of application in recognition systems, achieving state-of-the-art results.

show abstract

“…However, recognizing user's activities is still a tough task. Activity context recognition aims to infer a user's behavior from observations such as sensor data (Hu et al, 2011), and has various applications including medical care (Pollack et al, 2003), logistics service (Lin, 2006), robot soccer (Vail et al, 2007), plan recognition (Geib et al, 2008), etc. Many researchers make great attempts to find the most efficient way to recognize user's activities.…”

Section: Context and Activity Context Recognitionmentioning

confidence: 99%