Chengxiong Ruan

Learn More
Turn-by-turn navigation is a useful paradigm for assisting people with visual impairments during mobility as it reduces the cognitive load of having to simultaneously sense, localize and plan. To realize such a system, it is necessary to be able to automatically localize the user with sufficient accuracy, provide timely and efficient instructions and have(More)
Determining what features are important for face representation is quite challenging in Face Recognition. Real Adaboost performs remarkably in training classifiers for object detection which is a binary classification problem. As for Face Recognition, we should transform the multi-class problem into a binary one. In this paper, a feature selection method(More)
Most user studies of navigation applications for people with visual impairments have been limited by existing localization technologies, and appropriate instruction types and information needs have been determined through interviews. Using Wizard-of- Oz navigation interfaces, we explored how people with visual impairments respond to different instruction(More)
This paper investigates the local Gabor filters for facial expression recognition. The local Gabor filters can not only reduce the memory requirement but also lower the time of extracting Gabor features. By contracting among the recognition rates using different local Gabor filters, we can get that not each of the Gabor filters which choose different scales(More)
Automatic 3D facial expression recognition is still a challenging problem. This paper proposes the variation faces combining SVM to classify 3D facial expressions automatically as the flow of generating variation faces is without any manual intervention. To validate this strategy, the Fourier spectrum feature is explored and its highest recognition rate,(More)
Automatic crack detection on train wheel surface is highly crucial for safety and labor-cost efficiency. Algorithms have been explored to cope with the image based crack detection problem, but mostly for concrete structure or pavement surface. It's a tougher problem on train wheel surface because of noise such as scratch, rust, dirt, shadow and son on which(More)
This paper employs the interval weighted multi-template matching approach for accurate measurement. Generally, a vernier caliper is used as the subject to capture the analyzing image. However, the vernier caliper has many similar regions which all effect the measure accuracy. The traditional template matching with only one template is not suitable, as the(More)
Determining what features are optimal for face representation is quite a challenge in Face Recognition. Joint Boosting is a strong algorithm to select important features from feature pool. And, it outperforms other feature selection methods. But it takes quite a long time to train on a large training set. In this paper, an Accelerated Joint Boosting is(More)
  • 1