alicevision/popsift
A GPU-accelerated implementation of the SIFT (Scale-Invariant Feature Transform) algorithm using CUDA.
Not currently ranked — collecting fresh signals.
star history
PopSift implements the classic SIFT feature detection and description algorithm on NVIDIA GPUs via CUDA. It adheres closely to David Lowe’s original paper while achieving real-time performance on modern GPUs like the GTX 980 Ti. The library provides the core libpopsift artifact along with example applications and supports compute capabilities 3.5 and above.
Frequently asked
- What is alicevision/popsift?
- A GPU-accelerated implementation of the SIFT (Scale-Invariant Feature Transform) algorithm using CUDA.
- Is popsift open source?
- Yes — alicevision/popsift is open source, released under the MPL-2.0 license.
- What language is popsift written in?
- alicevision/popsift is primarily written in Cuda.
- How popular is popsift?
- alicevision/popsift has 497 stars on GitHub.
- Where can I find popsift?
- alicevision/popsift is on GitHub at https://github.com/alicevision/popsift.