Chaoyi Pang

Online human gesture recognition from motion data streams

Online human gesture recognition has a wide range of applications in computer vision, especially in human-computer interaction applications. Recent introduction of cost-effective depth cameras brings on a new trend of research on body-movement gesture recognition. However, there are two major challenges: i) how to continuously recognize gestures from unsegmented streams, and ii) how to differentiate different styles of a same gesture from other types of gestures. In this paper, we solve these two problems with a new effective and efficient feature extraction method that uses a dynamic matching approach to construct a feature vector for each frame and improves sensitivity to the features of different gestures and decreases sensitivity to the features of gestures within the same class. Our comprehensive experiments on MSRC-12 Kinect Gesture and MSR-Action3D datasets have demonstrated a superior performance than the stat-of-the-art approaches.

Action recognition by exploring data distribution and feature correlation

Human action recognition in videos draws strong research interest in computer vision because of its promising applications for video surveillance, video annotation, interactive gaming, etc. However, the amount of video data containing human actions is increasing exponentially, which makes the management of these resources a challenging task. Given a database with huge volumes of unlabeled videos, it is prohibitive to manually assign specific action types to these videos. Considering that it is much easier to obtain a small number of labeled videos, a practical solution for organizing them is to build a mechanism which is able to conduct action annotation automatically by leveraging the limited labeled videos. Motivated by this intuition, we propose an automatic video annotation algorithm by integrating semi-supervised learning and shared structure analysis into a joint framework for human action recognition. We apply our algorithm on both synthetic and realistic video datasets, including KTH [20], CareMedia dataset [1], Youtube action [12] and its extended version, UCF50 [2]. Extensive experiments demonstrate that the proposed algorithm outperforms the compared algorithms for action recognition. Most notably, our method has a very distinct advantage over other compared algorithms when we have only a few labeled samples.

Privacy-Preserving Fuzzy Matching Using a Public Reference Table

In this paper we address the problem of matching data from different databases using a third party, where the actual data can not be disclosed. The aim is to provide a mechanism for improved matching results across databases while preserving the privacy of sensitive information in those databases. This is particularly relevant with health related databases, where bringing data about patients together from multiple databases allows for important medical research, but the sensitive nature of the data requires that identifying information never be disclosed.

Dominating sets in directed graphs

We consider the problem of incrementally computing a minimal dominating set of a directed graph after the insertion or deletion of a set of arcs. Earlier results have either focused on the study of the properties that minimum (not minimal) dominating sets preserved or lacked to investigate which update affects a minimal dominating set and in what ways. In this paper, we first show how to incrementally compute a minimal dominating set on arc insertions. We then reduce the case of computing a minimal dominating set on arc deletions to the case of insertions. Some properties on minimal dominating sets are provided to support the incremental strategy. Lastly, we give a new bound on the size of minimum dominating sets based on those results.

Human action recognition based on semi-supervised discriminant analysis with global constraint

Human action recognition is an important area in computer vision and pattern recognition. Human joint position data are regarded as the most effective feature for this task. Depth camera using fringe projection techniques and related software provides us the capability to generate a large amount of human joint position data. However, these data cannot be used as the training data for supervised learning before the action labels are given, and manually labeling all the data is quite time-consuming. In this paper, we propose a novel algorithm named semi-supervised discriminant analysis with global constraint (SDG) which can better estimate the data distribution with both insufficient labeled data and sufficient unlabeled data. We use public mocap dataset HumanEva which is obtained by marker-based motion capture system, and our proposed skeleton dataset captured by depth camera for the evaluation. Experimental results demonstrate the effectiveness of our algorithm.

Semi-supervised Learning for Cyberbullying Detection in Social Networks

Current approaches on cyberbullying detection are mostly static: they are unable to handle noisy, imbalanced or streaming data efficiently. Existing studies on cyberbullying detection are mainly supervised learning approaches, assuming data is sufficiently pre-labelled. However this is impractical in the real-world situation where only a small number of labels are available in streaming data. In this paper, we propose a semi-supervised leaning approach that will augment training data samples and apply a fuzzy SVM algorithm. The augmented training technique automatically extracts and enlarges training set from the unlabelled streaming text, while learning is conducted by utilising a very small training set provided as an initial input. The experimental results indicate that the proposed augmented approach outperformed all other methods, and is suitable in the real-world situations, where sufficiently labelled instances are not available for training. For the proposed fuzzy SVM approach we handle complex and multidimensional data generated by streaming text, where the importance of features are discriminated for the decision function. The evaluation conducted on different experimental scenarios indicates the superiority of the proposed fuzzy SVM against all other methods.

Sentiment analysis for effective detection of cyber bullying

The rapid growth of social networking and gaming sites is associated with an increase of online bullying activities which, in the worst scenario, result in suicidal attempts by the victims. In this paper, we propose an effective technique to detect and rank the most influential persons (predators and victims). It simplifies the network communication problem through a proposed detection graph model. The experimental results indicate that this technique is highly accurate.

Structured Streaming Skeleton -- A New Feature for Online Human Gesture Recognition

Online human gesture recognition has a wide range of applications in computer vision, especially in human-computer interaction applications. The recent introduction of cost-effective depth cameras brings a new trend of research on body-movement gesture recognition. However, there are two major challenges: (i) how to continuously detect gestures from unsegmented streams, and (ii) how to differentiate different styles of the same gesture from other types of gestures. In this article, we solve these two problems with a new effective and efficient feature extraction method—Structured Streaming Skeleton (SSS)—which uses a dynamic matching approach to construct a feature vector for each frame. Our comprehensive experiments on MSRC-12 Kinect Gesture, Huawei/3DLife-2013, and MSR-Action3D datasets have demonstrated superior performances than the state-of-the-art approaches. We also demonstrate model selection based on the proposed SSS feature, where the classifier of squared loss regression with l2,1 norm regularization is a recommended classifier for best performance.

Unrestricted wavelet synopses under maximum error bound

Constructing Haar wavelet synopses under a given approximation error has many real world applications. In this paper, we take a novel approach towards constructing unrestricted Haar wavelet synopses under an error bound on uniform norm (L∞). We provide two approximation algorithms which both have linear time complexity and a (log N)-approximation ratio. The space complexities of these two algorithms are O (log N) and O (N) respectively. These two algorithms have the advantage of being both simple in structure and naturally adaptable for stream data processing. Unlike traditional approaches for synopses construction that rely heavily on examining wavelet coefficients and their summations, the proposed construction methods solely depend on examining the original data and are extendable to other findings. Extensive experiments indicate that these techniques are highly practical and surpass related ones in both efficiency and effectiveness.

Local correlation detection with linearity enhancement in streaming data

This paper addresses the challenges in detecting the potential correlation between numerical data streams, which facilitates the research of data stream mining and pattern discovery. We focus on local correlation with delay, which may occur in burst at different time in different streams, and last for a limited period. The uncertainty on the correlation occurrence and the time delay make it difficult to monitor the correlation online. Furthermore, the conventional correlation measure lacks the ability of reflecting visual linearity, which is more desirable in reality. This paper proposes effective methods to continuously detect the correlation between data streams. Our approach is based on the Discrete Fourier Transform to make rapid cross-correlation calculation with time delay allowed. In addition, we introduce a shape-based similarity measure into the framework, which refines the results by representative trend patterns to enhance the significance of linearity. The similarity of proposed linear representations can quickly estimate the correlation, and the window sliding strategy in segment level improves the efficiency for online detection. The empirical study demonstrates the accuracy of our detection approach, as well as more than