A CUDA implementation replaces FAST's branch-heavy tests with 32-bit bit masks and uses a semi-separable Sobel with circular buffers, reporting 2.2-4.5x faster FAST detection and up to 13x faster Harris scoring on embedded GPUs versus CUDA_ORB.
Features from accelerated segment test (fast)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Faster than Fast: Accelerating Oriented FAST Feature Detection on Low-end Embedded GPUs
A CUDA implementation replaces FAST's branch-heavy tests with 32-bit bit masks and uses a semi-separable Sobel with circular buffers, reporting 2.2-4.5x faster FAST detection and up to 13x faster Harris scoring on embedded GPUs versus CUDA_ORB.