I need cutting edge CV algorithms implemented that uses TBB, IPP and CUDA 7.0 for a security system to perform the following:
1) facial features extraction,
2) face recognition,
3) eye gaze detection
4) near-duplicates detection
5) licence plate recognition
6) Image mosaicing
7) camera calibration
and would also like to know what kind of additional Video analytics for surveillance systems have you implemented that you would recommend incorporating into the system.
To support real-time analysis of the above features for very high resolution feeds the code will be running on AWS large GPU instance with 32 CPU cores and Quad Nvidia-Grid GPUs (1500 cores each) and 64 GB of RAM. So use of IPP and CUDA7.0 to accelerate the performance is essential. The platform will be running on Ubuntu 14.04. If it makes it easier I can give you a prebuilt AWS image running OpenCV3.0, CUDA 7.0, IPP, TBB.
Input files should be of mp4 type with H.264 or H.265 encoding. Please let me know what additional details you might need, what is your process and pricing preferably broken down by functionality.
i have experiences in license plate recognition and face recognition
i am an expert in matlab and c++, opencv+cuda
i can do ur job.
Thanks for ur interest