Intel® Integrated Performance Primitives
Community support and discussions relating to developing high-performance vision, signal, security, and storage applications.
This community is designed for sharing of public information. Please do not share Intel or third-party confidential information here.

Fast nearest neighbor computation


Dear All,

I am looking for good c++ libraries for fast computation of the nearest neighbor on relatevly large dataset (100000 entries) and relatively high dimension (10-20 dimensions). The number of queries will be as large as the point dataset.

Do anyone know if such libraries are included in the IPP?

If not, may anyone redirect me to good libraries?

I thank you for your kind assistance.

My best regards, TRD

P.S. As I am new in this forum, I am not sure this question fits with this subsection. In that case, I would kindly ask to be redirected to the good place.

0 Kudos
1 Reply
Others may have better suggestions, but in general IPP is a signal processing (1D) and image processing (2D) library. For example, IPP has a nearest neighbor image resize algorithm which simply fills each point in the destination image with the single value at the "nearest" 2D coordinate in the source image. A quick google search shows the 'ANN' library for nearest neighbor search with high dimensionality. I have not used it, but is this the kind of functionality you're looking for?