Incremental Learning in Human Action Recognition Based on Snippets

Minhas, Rashid; Mohammed, Abdul Adeel; Wu, Qi

doi:10.1109/tcsvt.2011.2177182

Cited by 79 publications

(41 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Foreground masks, bounding boxes and action annotation are provided as ground truth. In this dataset, all the [21,37,40]. It can be seen that our approach achieves performance comparable to that of the traditional high-latency method for different action classes.…”

Section: Uiuc Datasetmentioning

confidence: 83%

“…Yang et al [41] introduced a patch based motion descriptor and presented a method to learn a transferable distance function for action recognition from a single video clip. Minhas et al [21] propose an adaptive PHOG features for action recognition using a small collection of video frames. Rose et al [24] propose an incremental nonparametric prediction system and combine it with sets of frame-by-frame local feature descriptors for online action recognition.…”

Section: Related Workmentioning

confidence: 99%

“…Same "clapping" clips are misclassified as "waving" since both classes alternate arms at a similar rate and in some clips both actions exhibit similar appearance as well. Figure 8 shows the per-class classification accuracy comparison on the KTH dataset vs. [21,37,40]. We choose these three works as they provide the per-class classification accuracy on the KTH dataset in their papers.…”

Section: Kth Datasetmentioning

confidence: 99%

See 2 more Smart Citations

Action recognition with low observational latency via part movement model

Liu

Wang

2016

Multimed Tools Appl

View full text Add to dashboard Cite

In this paper, we address the issue of recognizing human actions with low observational latency, which is vital for many applications such as virtual reality and interactive entertainment. Then our purpose is to achieve competitive action recognition performance from very short video clips. Such a task essentially is challenging because only very limited information is provided. To this end, we first develop a feature extraction method to exploit both motion (local flow) and appearance (local shape) features such that the information insufficiency can be effectively mitigated. Then we propose an action representation method named Part Movement Model (PMM), which explicitly captures the spatial-temporal structure of human actions and divides the actions into discriminative part movements. Consequently, the actions can be better represented and the competitive performance can be achieved although only short clips are used. Finally, we experimentally verify the effectiveness of the proposed method on three benchmark datasets. The results show that short clips of 6 − 7 frames (0.2 − 0.3 second video) are enough to achieve the recognition performance comparable to the baselines with high latency.

show abstract

Section: Uiuc Datasetmentioning

confidence: 83%

Section: Related Workmentioning

confidence: 99%

Section: Kth Datasetmentioning

confidence: 99%

See 1 more Smart Citation

Action recognition with low observational latency via part movement model

Liu

Wang

2016

Multimed Tools Appl

View full text Add to dashboard Cite

show abstract

“…The method achieves similar activity classification rates as other non-incremental approaches. In [29], snippet-level action recognition is targeted using a recursively trained classifier based on a single-hidden layer feed forward neural network, which is extended to present an incremental behaviour. Nonetheless, it also is adaptive, as the input weights are set randomly initially and then adjusted by means of a generalised inverse operation of the hidden layer weight matrices.…”

Section: B Incremental and Adaptive Human Action Recognitionmentioning

confidence: 99%

Adaptive Human Action Recognition With an Evolving Bag of Key Poses

Chaaraoui

Flórez-Revuelta

2014

IEEE Trans. Auton. Mental Dev.

View full text Add to dashboard Cite

Abstract-Vision-based human action recognition allows to detect and understand meaningful human motion. This makes it possible to perform advanced human-computer interaction, among other applications. In dynamic environments, adaptive methods are required to support changing scenario characteristics. Specifically, in human-robot interaction, smooth interaction between humans and robots can only be performed if these are able to evolve and adapt to the changing nature of the scenarios. In this paper, an adaptive vision-based human action recognition method is proposed. By means of an evolutionary optimisation method, adaptive and incremental learning of human actions is supported. Through an evolving bag of key poses, which models the learnt actions over time, the current learning memory is developed to recognise increasingly more actions or actors. The evolutionary method selects the optimal subset of training instances, features and parameter values for each learning phase, and handles the evolution of the model. The experimentation shows that our proposal achieves to adapt to new actions or actors successfully, by rearranging the learnt model. Stable and accurate results have been obtained on four publicly available RGB and RGB-D datasets, unveiling the method's robustness and applicability.

show abstract

“…Wang 3 -network) to meet the challenge of the so-called big data [15]. In addition, ELM has been put into diverse applications such as speaker recognition [16], neuroimage data classification [17], security assessment [18], data privacy [19], EEG and seizure detection [20], image quality assessment [21], image super-resolution [22], FPGA [23], face recognition [24], and human action recognition [25].…”

Section: Introductionmentioning

confidence: 99%