This paper treats tracking as a foreground/background classification problem and proposes anonline semisupervised learning framework. Initialized with a small number of labeled samples, semi-supervised learning treats each new sample as unlabeled data. Classification of new data and updating of the classifier are achieved simultaneously in a co-training framework. The object is represented using independent features and an online support vector machine (SVM) is built for each feature. The predictions from different features are fused by combining the confidence map from each classifier using a classifier weighting method which creates a final classifier that performs better than any classifier based on a single feature. The semi-supervised learning approach then uses the output of the combined confidence map to generate new samples and update the SVMs online. With this approach, the tracker gains increasing knowledge of the object and background and continually improves itself over time. Compared to other discriminative trackers, the online semi-supervised learning approach improves each individual classifier using the information from other features, thus leading to a more robust tracker. Experiments show that this framework performs better than state-of-the-art tracking algorithms on challenging sequences. 978-1-4244-1631-8/07/$25.00 ©2007 IEEE

Figure 1: Our edit propagation algorithm can be applied for various applications including color manipulation and matting. From left to right, these are the results of video recoloring, video color theme editing, image colorization and matting. Shown on the top are the input images (with a few user edits). Below are their corresponding output. AbstractWe propose a novel edit propagation algorithm for interactive image and video manipulations. Our approach uses the locally linear embedding (LLE) to represent each pixel as a linear combination of its neighbors in a feature space. While previous methods require similar pixels to have similar results, we seek to maintain the manifold structure formed by all pixels in the feature space. Specifically, we require each pixel to be the same linear combination of its neighbors in the result. Compared with previous methods, our proposed algorithm is more robust to color blending in the input data. Furthermore, since every pixel is only related to a few nearest neighbors, our algorithm easily achieves good runtime efficiency. We demonstrate our manifold preserving edit propagation on various applications.

The Earth Mover's Distance (EMD) is a similarity measure that captures perceptual difference between two distributions. Its computational complexity, however, prevents a direct use in many applications. This paper proposes a novel Differential EMD (DEMD) algorithm based on the sensitivity analysis of the simplex method and offers a speedup at orders of magnitude compared with its brute-force counterparts. The DEMD algorithm is discussed and empirically verified in the visual tracking context. The deformations of the distributions for objects at different time instances are accommodated well by the EMD, and the differential algorithm makes the use of EMD in real-time tracking possible. To further reduce the computation, signatures, i.e., variable-size descriptions of distributions, are employed as an object representation. The new algorithm models and estimates local background scenes as well as foreground objects to handle scale changes in a principled way. Extensive quantitative evaluation of the proposed algorithm has been carried out using benchmark sequences and the improvement over the standard Mean Shift tracker is demonstrated.

For the task of mobility analysis of 3D shapes, we propose joint analysis for simultaneous motion part segmentation and motion attribute estimation, taking a single 3D model as input. The problem is significantly different from those tackled in the existing works which assume the availability of either a pre-existing shape segmentation or multiple 3D models in different motion states. To that end, we develop Shape2Motion which takes a single 3D point cloud as input, and jointly computes a mobility-oriented segmentation and the associated motion attributes. Shape2Motion is comprised of two deep neural networks designed for mobility proposal generation and mobility optimization, respectively. The key contribution of these networks is the novel motion-driven features and losses used in both motion part segmentation and motion attribute estimation. This is based on the observation that the movement of a functional part preserves the shape structure. We evaluate Shape2Motion with a newly proposed benchmark for mobility analysis of 3D shapes. Results demonstrate that our method achieves the state-of-the-art performance both in terms of motion part segmentation and motion attribute estimation.

In this paper we propose a novel alpha matting method with local and nonlocal smooth priors. We observe that the manifold preserving editing propagation [4] essentially introduced a nonlocal smooth prior on the alpha matte. This nonlocal smooth prior and the well known local smooth prior from matting Laplacian complement each other. So we combine them with a simple data term from color sampling in a graph model for nature image matting. Our method has a closed-form solution and can be solved efficiently. Compared with the state-of-the-art methods, our method produces more accurate results according to the evaluation on standard benchmark datasets.

No abstract

No abstract

Detecting the sources or destinations that have communicated with a large number of distinct destinations or sources during a small time interval is an important problem in network measurement and security. Previous detection approaches are not able to deliver the desired accuracy at high link speeds (10 to 40 Gbps). In this work, we propose two novel algorithms that provide accurate and efficient solutions to this problem. Their designs are based on the insight that sampling and data streaming are often suitable for capturing different and complementary regions of the information spectrum, and a close collaboration between them is an excellent way to recover the complete information. Our first solution builds on the standard hash-based flow sampling algorithm. Its main innovation is that the sampled traffic is further filtered by a data streaming module which allows for much higher sampling rate and hence much higher accuracy. Our second solution is more sophisticated but offers higher accuracy. It combines the power of data streaming in efficiently estimating quantities associated with a given identity, and the power of sampling in collecting a list of candidate identities. The performance of both solutions are evaluated using both mathematical analysis and trace-driven experiments on real-world Internet traffic.