Zhiqiang Wei

A Review on Human Activity Recognition Using Vision-Based Method

Human activity recognition (HAR) aims to recognize activities from a series of observations on the actions of subjects and the environmental conditions. The vision-based HAR research is the basis of many applications including video surveillance, health care, and human-computer interaction (HCI). This review highlights the advances of state-of-the-art activity recognition approaches, especially for the activity representation and classification methods. For the representation methods, we sort out a chronological research trajectory from global representations to local representations, and recent depth-based representations. For the classification methods, we conform to the categorization of template-based methods, discriminative models, and generative models and review several prevalent methods. Next, representative and available datasets are introduced. Aiming to provide an overview of those methods and a convenient way of comparing them, we classify existing literatures with a detailed taxonomy including representation and classification methods, as well as the datasets they used. Finally, we investigate the directions for future research.

Robust skin detection in real-world images

Proposed scheme incorporates color property, texture property and region property.Robust skin seeds are acquired by combining color property and texture property.SCTGC incorporates color property, texture property and region property. Human skin detection in images is desirable in many practical applications, e.g., human-computer interaction and adult-content filtering. However, existing methods are mainly suffer from confusing backgrounds in real-world images. In this paper, we try to address this issue by exploring and combining several human skin properties, i.e. color property, texture property and region property. First, images are divided into superpixels, and robust skin seeds and background seeds are acquired through color property and texture property of skin. Then we combining color, region and texture properties of skin by proposing a novel skin color and texture based graph cuts (SCTGC) to acquire the final skin detection results. Comprehensive and comparative experiments show that the proposed method achieves promising performance and outperforms many state-of-the-art methods over publicly available challenging datasets with a great part of hard images.

How Your Portrait Impresses People?: Inferring Personality Impressions from Portrait Contents

Whenever looking at a strangers portrait, besides observable appearance, we always build a personality impression implicitly in our subconscious. It is quite interesting to ask how a portrait impresses people. This paper presents a novel method to infer personality impression from portrait. Firstly, a questionnaire is applied to demonstrate the consistence of peoples impression. And then personality-related features are explored through the statistical analysis method. Finally, features are trained using Support Vector Machine. Experimental results demonstrate our method could achieve a precision of 52.14% and a recall of 52.78% on inferring 4 personalities from 2,463 randomly selected portraits of people downloaded from Google images. Improvements of 44.04% and 37.91% are reported compared to a baseline method. And features contribution analysis deeply unveils the correspondence between portrait contents and personality impressions. Demonstrations with respect to visual patterns in portrait collages of different personalities further prove the effectiveness of the proposed method. Furthermore, we apply our method to analyze portraits of Hillary Clinton and obtain an interesting multifaceted figure of this famous politics, which is another proof of both our concept and method.

Is Your First Impression Reliable? Trustworthy Analysis Using Facial Traits in Portraits

As a basic human quality, trustworthiness plays an important role in social communications. In this paper, we proposed a novel concept to predict people’s trustworthiness at first sight using facial traits. Firstly, personality-toward traits were designed from psychology, including permanent traits and transient traits. Then, a mixture of feature descriptors consisting of Histogram of Gradients (HOG), Local Binary Patterns (LBP) and geometrical descriptions were adopted to describe personality traits. Finally, we trained the personality traits by LibSVM to determine trustworthiness of a person using portrait. Experiments demonstrated the effectiveness of our method by improving the precision by 33.60%, recall by 20.33% and F1-measure by 25.63% when determining whether a person is trustworthy or not comparing to a baseline method. Feature contribution analysis was applied to deeply unveil the correspondence between features and personality. Demonstration showed visual patterns in portrait collages of trustworthy people that further proved effectiveness of our method.

Structured Light Encoding Research Based on Sub-Pixel Edge Detection

Coded structured light is an effective technique to acquiring shape in stereovision. Coding strategy is one of the core techniques because it effects the measure precision and efficiency directly. There are two major group of techniques in color structured light encoding. One is to define coloured multi-slit, which is suitable for locating intensity peaks in the image. The other is to define coloured stripe patterns, which is suitable to locate edges. In this paper, we combine the two methods to design new coloured stripe patterns, so that both intensity peaks and edges can be located. Then, edge detection method with sub-pixel accuracy is performed and the detected stripes centres and the edges between adjacent stripes are matched with the projected pattern, which is the foundation for further dense reconstructions. The experimental results show that our methods are very powerful in terms of accuracy and resolution. Keywordsshape acquisition; color structured light; sub-pixel detection

Human body segmentation based on shape constraint

Human body segmentation is essential for many practical applications, e.g., video surveillance analysis in intelligent urban. However, existing methods mainly suffer from various human poses. In this paper, we try to address this issue by introducing human shape constraint. First, human pose estimation is performed, and locations of human body parts are determined. Contrast to the previous work, we just use the human body parts with high precision. Then we combines the star convexity and the human body parts’ locations as shape constraint. The final segmentation results are acquired through the optimization step. Comprehensive and comparative experimental results demonstrate that the proposed method achieves promising performance and outperforms many state-of-the-art methods over publicly available challenging datasets.

Social Media Profiler: Inferring Your Social Media Personality from Visual Attributes in Portrait

In this paper, we introduce an interesting but challenging problem: how to infer social media personality from portrait. To address this problem, we jointly consider social media content and behavior information. Specifically, first, we represent social media personality as a reflection in accordance with user behaviors in social media. Second, by means of clustering, people are divided into eight groups and labeled with different personality types. Upon regression analysis, discriminative visual attributes for personality classification are determined. Third, low-level features of selected visual attributes are trained to predict personality from given portrait. To evaluate the proposed method, we collect images of people from the internet and the behaviors of these people from their micro-blog. Comprehensive experiments demonstrate that the proposed method can achieve significant performance gain over the existing method.

Indirect human activity recognition based on optical flow method

A new method to recognize human activity with videos from a wearable camera is proposed in this paper. With a camera mounted to a human body, the moving subject wont appear in the video when the person is in some motions. But we can estimate the activity from the changes of scenes in videos. Optical flow method is a common method to calculate motion vectors of objects in two adjacent images. For higher precision, in this paper, we use Lucas-Kanade optical flow method with pyramid structure to calculate the optical flow of scenes which can reflect peoples motion to some extents. When key information is extracted from the optical flow field, we design a feature descriptor to describe the motion in frames in a video. The feature descriptor contains angels, bounce information and other important information which can distinguish different motion. After getting feature descriptors, we use support vector machine to classify different motions with a machine learning method. Experimental results show that our method successfully identifies motion such as walking, running, going upstairs and going downstairs. Compared with methods based on blocking-matching, this method has fewer costs and has higher precision.

Pedestrian Detection Based on a New Two-Step Framework

In this paper, we propose a new framework in pedestrian detection using a two-step classification algorithm, which is a “coarse to fine” course. The framework consists of a full-body detection (FBD) step and a head-shoulder detection (HSD) step. The FBD step uses fusion of Haar-like and HOG features to get better performance, and the HSD step utilizes edgelet features for classification and detection. The pedestrian data is obtained from MIT, INRIA dataset and surveillance videos for training. The experiment carried out on videos from campus and CAVIAR dataset illustrates that the proposed method is robust and feasible enough for pedestrian detection and could handle occlusions more accurately than other methods.

A Novel Particle Filter Method for Mobile Robot Localization

Particle filter is a powerful tool for mobile robot localization based on Sequential Monte Carlo framework. However, it needs a large number of samples to properly approximate the posterior density of the state evolution, which makes it computational expensive. In this paper, an improved particle filter is proposed by adopting an EKF proposal distribution and Support Vector Regression (SVR). The proposed particle filter uses an EKF proposal to provide good quality samples, and an SVR based re-weighting scheme to re-weight the sample more accurately. Thus the effectiveness and diversity of samples are maintained meanwhile impoverishment is avoided as much as possible. Experiment results show that the proposed particle filter can work with a small sample set effectively and is more precise for mobile robot localization than classical particle filter.