Location:
Search - cuda m
Search list
Description: md5_cuda编程,主要是针对GPU加速-md5_cuda programming is targeted at the GPU to accelerate the
Platform: |
Size: 31744 |
Author: 张宏 |
Hits:
Description: 如何在MFC中调用CUDA,环境: Windows Vista SP1,Microsoft Visual Studio 2005,CUDA 2.0
步骤:1.创建一个对话框的MFC程序
-How to call MFC CUDA, Environment: Windows Vista SP1, Microsoft Visual Studio 2005, CUDA 2.0: 1. Create a MFC dialog box program
Platform: |
Size: 642048 |
Author: zhouxiaol |
Hits:
Description: Robust Non-negative Dictionary Learning for Visual Tracking
The provided codes could be either embedded into the benchmark framework of paper Online Object Tracking: A Benchmark (CVPR2013) (You can find details here: http://visual-tracking.net/) or run on individual sequence.
To run the benchmark, just put the entire folder into the /trackers folder in the benchmark code base, and modify the configTrackers.m in util folder. DLT gets an AUC of 0.436, which ranks 5th among 26 in the benchmark by 19/03/2014. We don t tune parameters for single sequence in this case, all the parameters are stored in trackparam_DLT.m.
To run on individual video, you need to modify the dataPath and title in run_individual.m.
If you run MATLAB version after 2012, and have a CUDA compatible GPU installed, you may enjoy the fast computation speed by GPU, just set useGPU to true in trackparam_DLT.m and run_individual.m!
-Robust Non-negative Dictionary Learning for Visual Tracking
The provided codes could be either embedded into the benchmark framework of paper Online Object Tracking: A Benchmark (CVPR2013) (You can find details here: http://visual-tracking.net/) or run on individual sequence.
To run the benchmark, just put the entire folder into the /trackers folder in the benchmark code base, and modify the configTrackers.m in util folder. DLT gets an AUC of 0.436, which ranks 5th among 26 in the benchmark by 19/03/2014. We don t tune parameters for single sequence in this case, all the parameters are stored in trackparam_DLT.m.
To run on individual video, you need to modify the dataPath and title in run_individual.m.
If you run MATLAB version after 2012, and have a CUDA compatible GPU installed, you may enjoy the fast computation speed by GPU, just set useGPU to true in trackparam_DLT.m and run_individual.m!
Platform: |
Size: 22211584 |
Author: mohit |
Hits: