SIFT
A classical local feature description algorithm in computer vision, proposed by David Lowe in 1999. SIFT detects stable keypoints in scale space and generates 128-dimensional descriptor vectors, providing invariance to rotation and illumination changes. It became a foundational method in image matching, object recognition, and related computer vision tasks.
Timeline (last 90 days)
SIFT (Scale-Invariant Feature Transform) was proposed by David Lowe in 1999, detecting stable keypoints across different scale spaces and generating 128-dimensional descriptor vectors
The patent for SIFT expired in 2020
ORB (Oriented FAST and Rotated BRIEF) is fully open-source, approximately 10–100 times faster than SIFT, and suitable for real-time scenarios
Earlier functional prediction tools such as CADD, PolyPhen-2, and SIFT can also score large numbers of variants, but AlphaGenome has the ability to predict multi-dimensional molecular phenotypes from sequence.
Traditional keypoint detection methods include SIFT and ORB, while deep learning methods include SuperPoint.
Early variant prediction tools such as SIFT and PolyPhen primarily relied on evolutionary conservation and protein structure information, with accuracy rates of approximately 70–80%.
All Facts (6)
SIFT (Scale-Invariant Feature Transform) was proposed by David Lowe in 1999, detecting stable keypoints across different scale spaces and generating 128-dimensional descriptor vectors
50%UnverifiedThe patent for SIFT expired in 2020
50%UnverifiedORB (Oriented FAST and Rotated BRIEF) is fully open-source, approximately 10–100 times faster than SIFT, and suitable for real-time scenarios
50%UnverifiedEarlier functional prediction tools such as CADD, PolyPhen-2, and SIFT can also score large numbers of variants, but AlphaGenome has the ability to predict multi-dimensional molecular phenotypes from sequence.
50%UnverifiedTraditional keypoint detection methods include SIFT and ORB, while deep learning methods include SuperPoint.
50%UnverifiedEarly variant prediction tools such as SIFT and PolyPhen primarily relied on evolutionary conservation and protein structure information, with accuracy rates of approximately 70–80%.
50%