NVIDIA Introduces Open-Source Project to Accelerate Computer Vision Cloud Applications

CV-CUDA combines accelerated image pre- and post-processing algorithms and tools to process 10x the number of images at a single cost.

This results in higher throughput as well as lower cloud-computing expenses. 

CV-CUDA can process 10 times the amount of streams that can be processed using a single GPU.

This helps developers work more quickly when creating.

video content in 3D worlds and recommender systems based on images, image recognition, and video conference.

CV-CUDA improves the post- and pre-processing pipelines by hand-optimizing CUDA kernels.

is natively integrated with C/C++ Python and common Deep Learning frameworks such as PyTorch.