• Publications
  • Influence
Human Action Recognition Using DFT
  • S. Kumari, S. Mitra
  • Computer Science
  • Third National Conference on Computer Vision…
  • 15 December 2011
The proposed novel action recognition algorithm uses discrete Fourier transform (DFT) of the small image block to deal with the noise caused because of illumination effects, blurring, false contour etc. Expand
Parallelizing OPTICS for Commodity Clusters
The proposed DOPTICS algorithm is a parallelized version of a popular density based cluster-ordering algorithm OPTICS that is found to scale well with increasing number of processing elements and to achieve high parallelism. Expand
Exact, Fast and Scalable Parallel DBSCAN for Commodity Platforms
A grid-based DBSCAN algorithm, GridDBSCAN, is presented, which is significantly faster than the state-of-the-art sequential DBS CAN and its parallel implementations, and also proposes scalable parallel implementations of GridD BSCAN to leverage a multicore commodity cluster. Expand
A Fast, Scalable SLINK Algorithm for Commodity Cluster Computing Exploiting Spatial Locality
This paper presents a novel optimization of SLINK algorithm, GridSLINK, which is an order of magnitude faster than the existing state-of-the-art implementation and is benchmarked against the best existing parallel algorithm in literature and found to outperform the latter. Expand
Scalable Parallel Algorithms for Shared Nearest Neighbor Clustering
A new sequential SNN algorithm, R-SNN, which uses R-tree for executing neighborhood queries efficiently and exploiting spatial locality to minimize memory usage is presented, which is found up to 77 times faster when tested on various real datasets. Expand
μDBSCAN: An Exact Scalable DBSCAN Algorithm for Big Data Exploiting Spatial Locality
This work proposes a micro-cluster based DBSCAN algorithm, μDBSCAN, which identifies core-points even without performing neighbourhood queries and becomes instrumental in reducing the run-time of the algorithm, which significantly reduces the computation time per neighbourhood query while producing exact DBS CAN clusters. Expand
A Parallel Framework for Grid-Based Bottom-Up Subspace Clustering
An efficient parallel framework for grid-based bottom-up subspace clustering algorithms is developed, considering popular algorithms belonging to this category, and exhibits impressive speedup and scalability on real datasets. Expand
Spatial Locality Aware, Fast, and Scalable SLINK Algorithm for Commodity Clusters
Single linkage (SLINK) hierarchical clustering algorithm is a preferred clustering algorithm over traditional partitioning-based clustering as it does not require the number of clusters as input.Expand
Parallel SLINK for big data
A parallel SLINK algorithm, sGrid SLINK, is presented, a parallel algorithm which fully exploits a multi-core cluster system and is able to cluster 200 million data points in only 1317 s (less than 22 min), no existing parallel SLink algorithm is capable of such efficient clustering of Big Data. Expand
Parallelizing OPTICS for multicore systems
This work proposes a parallel version of OPTICS for shared memory multi-core systems using a master-slave pattern for parallelization and argues that this approach is well suited for dense datasets in particular. Expand