English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This work details a highly efficient implementation of the 3D scale-invariantfeature transform (SIFT) algorithm, for the purpose of machine learning fromlarge sets of volumetric medical image data. The primary operations of the 3DSIFT code are implemented on a graphics processing unit (GPU), includingconvolution, sub-sampling, and 4D peak detection from scale-space pyramids. Theperformance improvements are quantified in keypoint detection andimage-to-image matching experiments, using 3D MRI human brain volumes ofdifferent people. Computationally efficient 3D keypoint descriptors areproposed based on the Binary Robust Independent Elementary Feature (BRIEF)code, including a novel descriptor we call Ranked Robust Independent ElementaryFeatures (RRIEF), and compared to the original 3D SIFT-Rankmethod\citep{toews2013efficient}. The GPU implementation affords a speedup ofapproximately 7X beyond an optimised CPU implementation, where computation timeis reduced from 1.4 seconds to 0.2 seconds for 3D volumes of size (145, 174,145) voxels with approximately 3000 keypoints. Notable speedups include theconvolution operation (20X), 4D peak detection (3X), sub-sampling (3X), anddifference-of-Gaussian pyramid construction (2X). Efficient descriptors offer aspeedup of 2X and a memory savings of 6X compared to standard SIFT-Rankdescriptors, at a cost of reduced numbers of keypoint correspondences,revealing a trade-off between computational efficiency and algorithmicperformance. The speedups gained by our implementation will allow for a moreefficient analysis on larger data sets. Our optimized GPU implementation of the3D SIFT-Rank extractor is available athttps://github.com/CarluerJB/3D_SIFT_CUDA.

GPU optimization of the 3D Scale-invariant Feature Transform Algorithm and a Novel BRIEF-inspired 3D Fast Descriptor