A Split-and-Merge Dictionary Learning Algorithm for Sparse Representation: Application to Image Denoising

In big data image/video analytics, we encounter the problem of learning an over-complete dictionary for sparse representation from a large training dataset, which cannot be processed at once because of storage and computational constraints. To tackle the problem of dictionary learning in such scenarios, we propose an algorithm that exploits the inherent clustered structure of the training data and make use of a divide-and-conquer approach. The fundamental idea behind the algorithm is to partition the training dataset into smaller clusters, and learn local dictionaries for each cluster. Subsequently, the local dictionaries are merged to form a global dictionary. Merging is done by solving another dictionary learning problem on the atoms of the locally trained dictionaries. This algorithm is referred to as the split-and-merge algorithm. We show that the proposed algorithm is efficient in its usage of memory and computational complexity, and performs on par with the standard learning strategy, which operates on the entire data at a time. As an application, we consider the problem of image denoising. We present a comparative analysis of our algorithm with the standard learning techniques that use the entire database at a time, in terms of training and denoising performance. We observe that the split-and-merge algorithm results in a remarkable reduction of training time, without significantly affecting the denoising performance.

Sparse Representation based Anomaly Detection using HOMV in H.264 Compressed Videos

In this paper, we have proposed an anomaly detection algorithm based on Histogram of Oriented Motion Vectors (HOMV) 1] in sparse representation framework. Usual behavior is learned at each location by sparsely representing the HOMVs over learnt normal feature bases obtained using an online dictionary learning algorithm. In the end, anomaly is detected based on the likelihood of the occurrence of sparse coefficients at that location. The proposed approach is found to be robust compared to existing methods as demonstrated in the experiments on UCSD Ped1 and UCSD Ped2 datasets.

Status of the MSSM Higgs sector using global analysis and direct search bounds, and future prospects at the High Luminosity LHC

In this paper, we search for the regions of the phenomenological minimal supersymmetric standard model (pMSSM) parameter space where one can expect to have moderate Higgs mixing angle (alpha) with relatively light (up to 600 GeV) additional Higgses after satisfying the current LHC data. We perform a global fit analysis using most updated data (till December 2014) from the LHC and Tevatron experiments. The constraints coming from the precision measurements of the rare b-decays B-s -> mu(+)mu(-) and b -> s gamma are also considered. We find that low M-A(less than or similar to 350) and high tan beta(greater than or similar to 25) regions are disfavored by the combined effect of the global analysis and flavor data. However, regions with Higgs mixing angle alpha similar to 0.1-0.8 are still allowed by the current data. We then study the existing direct search bounds on the heavy scalar/pseudoscalar (H/A) and charged Higgs boson (H-+/-) masses and branchings at the LHC. It has been found that regions with low to moderate values of tan beta with light additional Higgses (mass <= 600 GeV) are unconstrained by the data, while the regions with tan beta > 20 are excluded considering the direct search bounds by the LHC-8 data. The possibility to probe the region with tan beta <= 20 at the high luminosity run of LHC are also discussed, giving special attention to the H -> hh, H/A -> t (t) over bar and H/A -> tau(+)tau(-) decay modes.

Sparse representation-based human detection: a scale-embedded dictionary approach

Human detection is a complex problem owing to the variable pose that they can adopt. Here, we address this problem in sparse representation framework with an overcomplete scale-embedded dictionary. Histogram of oriented gradient features extracted from the candidate image patches are sparsely represented by the dictionary that contain positive bases along with negative and trivial bases. The object is detected based on the proposed likelihood measure obtained from the distribution of these sparse coefficients. The likelihood is obtained as the ratio of contribution of positive bases to negative and trivial bases. The positive bases of the dictionary represent the object (human) at various scales. This enables us to detect the object at any scale in one shot and avoids multiple scanning at different scales. This significantly reduces the computational complexity of detection task. In addition to human detection, it also finds the scale at which the human is detected due to the scale-embedded structure of the dictionary.