Wedgelet Enhanced Appearance Models

Darkner, Sune; Larsen, Rasmus; Stegmann, Mikkel Bille; Ersbøll, Bjarne Kjær

doi:10.1109/cvpr.2004.478

Cited by 7 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, several improvements are proposed to achieve this aim. Some methods are proposed to reduce the dimension of the texture, such as the Haar wavelet [ 12 ], the wedgelet-based regression tree [ 13 ], and the local sampling [ 14 ]. However, these methods improve efficiency at the expense of decreasing accuracy or losing detail information.…”

Section: Introductionmentioning

confidence: 99%

Efficient Parallel Implementation of Active Appearance Model Fitting Algorithm on GPU

Wang

Zhu

et al. 2014

The Scientific World Journal

View full text Add to dashboard Cite

The active appearance model (AAM) is one of the most powerful model-based object detecting and tracking methods which has been widely used in various situations. However, the high-dimensional texture representation causes very time-consuming computations, which makes the AAM difficult to apply to real-time systems. The emergence of modern graphics processing units (GPUs) that feature a many-core, fine-grained parallel architecture provides new and promising solutions to overcome the computational challenge. In this paper, we propose an efficient parallel implementation of the AAM fitting algorithm on GPUs. Our design idea is fine grain parallelism in which we distribute the texture data of the AAM, in pixels, to thousands of parallel GPU threads for processing, which makes the algorithm fit better into the GPU architecture. We implement our algorithm using the compute unified device architecture (CUDA) on the Nvidia's GTX 650 GPU, which has the latest Kepler architecture. To compare the performance of our algorithm with different data sizes, we built sixteen face AAM models of different dimensional textures. The experiment results show that our parallel AAM fitting algorithm can achieve real-time performance for videos even on very high-dimensional textures.

show abstract

Section: Introductionmentioning

confidence: 99%

Efficient Parallel Implementation of Active Appearance Model Fitting Algorithm on GPU

Wang

Zhu

et al. 2014

The Scientific World Journal

View full text Add to dashboard Cite

show abstract

“…And then compressing schemes are used to reduce the data dimension. Darkner et al [15] propose to reduce the textural components based on wedgelet based regression tree and report a highest compression ratio of 1:150.…”

Section: Introductionmentioning

confidence: 99%

Gabor-based texture representation in AAMs

Gao

Tao

et al. 2008

2008 IEEE International Conference on Systems, Man and Cybernetics

View full text Add to dashboard Cite

Active Appearance Models (AAMs) are generative models which can describe deformable objects. However, the texture in basic AAMs is represented using intensity values. Despite its simplicity, this representation does not contain enough information for image matching. In this paper, we firstly propose to utilize Gabor filters to represent the image texture. The benefit of Gabor-based representation is that it can express local structures of an image. As a result, this representation can lead to more accurate matching when condition changes. Given the problem of the excessive storage and computational complexity of the Gabor, three different Gabor-based image representations are used in AAMs: (1) GaborD is the sum of Gabor filter responses over directions, (2) GaborS is the sum of Gabor filter responses over scales, and (3) GaborSD is the sum of Gabor filter responses over scales and directions. Through a large number of experiments, we show that the proposed Gabor representations lead to more accurate and reliable matching between model and images.

show abstract

“…In this case recently introduced wedgelets [8] are used with success in such tasks as compression [12,13,20,23], denoising [7,9,12,15], image segmentation [5,22] or edge detection [14], to mention a few. Unfortunately, the computation time of wedgelet transform, which is used in all these applications, is rather slow.…”

Section: Introductionmentioning

confidence: 99%

Moments-Based Fast Wedgelet Transform

Lisowska

2010

J Math Imaging Vis

View full text Add to dashboard Cite

In the paper the moments-based fast wedgelet transform has been presented. In order to perform the classical wedgelet transform one searches the whole wedgelets' dictionary to find the best matching. Whereas in the proposed method the parameters of wedgelet are computed directly from an image basing on moments computation. Such parameters describe wedgelet reflecting the edge present in the image. However, such wedgelet is not necessarily the best one in the meaning of Mean Square Error. So, to overcome that drawback, the method which improves the matching result has also been proposed. It works in the way that the better matching one needs to obtain the longer time it takes. The proposed transform works in linear time with respect to the number of pixels of the full quadtree decomposition of an image. More precisely, for an image of size N × N pixels the time complexity of the proposed wedgelet transform is O(N 2 log 2 N).

show abstract

Wedgelet Enhanced Appearance Models

Cited by 7 publications

References 31 publications

Efficient Parallel Implementation of Active Appearance Model Fitting Algorithm on GPU

Efficient Parallel Implementation of Active Appearance Model Fitting Algorithm on GPU

Gabor-based texture representation in AAMs

Moments-Based Fast Wedgelet Transform

Contact Info

Product

Resources

About