I work for Synesis (http://synesis.ru/) which makes solution for video analytics (motion detection, face recognition, license plate's recognition, and many other algorithms). Early we used standard algorithms including algorithms from OpenCV. But performance of these algorithms is not very good. And I was forced to perform optimization of the algorithms with using of different SIMD instruction. My optimized code is 9 times faster (on average). And I think that Simd Library can be usefull for other programers if they need of fast algorithms for computer vision.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Could the developer possibly provide some information on the motivation behind this package and how it compares to OpenCV?
I work for Synesis (http://synesis.ru/) which makes solution for video analytics (motion detection, face recognition, license plate's recognition, and many other algorithms). Early we used standard algorithms including algorithms from OpenCV. But performance of these algorithms is not very good. And I was forced to perform optimization of the algorithms with using of different SIMD instruction. My optimized code is 9 times faster (on average). And I think that Simd Library can be usefull for other programers if they need of fast algorithms for computer vision.