Feng Kang

Learn More
Many computer vision applications, such as scene analysis and medical image interpretation, are ill-suited for traditional classification where each image can only be associated with a single class. This has stimulated recent work in multi-label learning where a given image can be tagged with multiple class labels. A serious problem with existing approaches(More)
Spectral clustering enjoys its success in both data clustering and semi-supervised learning. But, most spectral clustering algorithms cannot handle multi-class clustering problems directly. Additional strategies are needed to extend spectral clustering algorithms to multi-class clustering problems. Furthermore, most spectral clustering algorithms employ(More)
The goal of automatic image annotation is to automatically generate annotations for images to describe their content. In the past, statistical machine translation models have been successfully applied to automatic image annotation task [8]. It views the process of annotating images as a process of translating the content from a 'visual language' to textual(More)
Cinnabar (HgS) is used in traditional medicines, and total Hg content is used for risk assessment of cinnabar-containing traditional medicines such as Zhu-Sha-An-Shen-Wan (ZSASW). Is ZSASW or cinnabar toxicologically similar to common mercurials? Adult Sprague-Dawley rats were gavaged with ZSASW (1.4 g/kg), cinnabar (0.2g/kg), HgCl(2) (0.02 g/kg), MeHg(More)
In this paper, we derive a new composition law obtained by substituting a B-series into the vector field appearing in another B-series. We derive explicit formulas for the computation of this law and study its algebraic properties. We then focus on the specific case of Hamiltonian vector fields. It is shown that this new law allows a convenient derivation(More)
Video annotation is an expensive but necessary task for most vision and learning problems that require building models of visual semantics. This annotation gets prohibitively expensive especially when annotation has to happen at finer grained levels of regions in the videos. One way around the finer grained annotation dilemma is to support annotation at(More)
Based on simulation software GT-drive, the author analyzed the transmission system performance of a passenger car with diesel engine and provided the appropriate research methods. Firstly, the numerical simulation model of a vehicle was built based on vehicle weight, frontal area, rolling, air-drag coefficient, etc. The different matching schemes were(More)
Statistical learning techniques provide a robust framework for learning representations of semantic concepts from multimedia features. The bottleneck is the number of training samples needed to construct robust models. This is particularly expensive when the annotation needs to happen at finer granularity. We present a novel approach where the annotations(More)