k-d tree

Abstract: An algorithm and data structure are presented for searching a file containing N records, each described by k real valued keys, for the m closest matches or nearest neighbors to a given query record. The computation required to organize the file is proportional to kNlogN. The expected number of records examined in each search is independent of the file size. The expected computation to perform each search is proportional to logN. Empirical evidence suggests that except for very small files, this algorithm is considerably faster than other methods.

Abstract: In this paper, we look at improving the KD-tree for a specific usage: indexing a large number of SIFT and other types of image descriptors. We have extended priority search, to priority search among multiple trees. By creating multiple KD-trees from the same data set and simultaneously searching among these trees, we have improved the KD-treepsilas search performance significantly.We have also exploited the structure in SIFT descriptors (or structure in any data set) to reduce the time spent in backtracking. By using Principal Component Analysis to align the principal axes of the data with the coordinate axes, we have further increased the KD-treepsilas search performance.

Abstract: We present an algorithm for constructing kd-trees on GPUs. This algorithm achieves real-time performance by exploiting the GPU's streaming architecture at all stages of kd-tree construction. Unlike previous parallel kd-tree algorithms, our method builds tree nodes completely in BFS (breadth-first search) order. We also develop a special strategy for large nodes at upper tree levels so as to further exploit the fine-grained parallelism of GPUs. For these nodes, we parallelize the computation over all geometric primitives instead of nodes at each level. Finally, in order to maintain kd-tree quality, we introduce novel schemes for fast evaluation of node split costs.As far as we know, ours is the first real-time kd-tree algorithm on the GPU. The kd-trees built by our algorithm are of comparable quality as those constructed by off-line CPU algorithms. In terms of speed, our algorithm is significantly faster than well-optimized single-core CPU algorithms and competitive with multi-core CPU algorithms. Our algorithm provides a general way for handling dynamic scenes on the GPU. We demonstrate the potential of our algorithm in applications involving dynamic scenes, including GPU ray tracing, interactive photon mapping, and point cloud modeling.

Abstract: This note presents a simplification and generalization of an algorithm for searchingk-dimensional trees for nearest neighbors reported by Friedmanet al [3]. If the distance between records is measured usingL2, the Euclidean norm, the data structure used by the algorithm to determine the bounds of the search space can be simplified to a single number. Moreover, because distance measurements inL2 are rotationally invariant, the algorithm can be generalized to allow a partition plane to have an arbitrary orientation, rather than insisting that it be perpendicular to a coordinate axis, as in the original algorithm. When ak-dimensional tree is built, this plane can be found from the principal eigenvector of the covariance matrix of the records to be partitioned. These techniques and others yield variants ofk-dimensional trees customized for specific applications. It is wrong to assume thatk-dimensional trees guarantee that a nearest-neighbor query completes in logarithmic expected time. For smallk, logarithmic behavior is observed on all but tiny trees. However, for largerk, logarithmic behavior is achievable only with extremely large numbers of records. Fork = 16, a search of ak-dimensional tree of 76,000 records examines almost every record.