I’m working on a project that requires running a hand-tracking module from Mediapipe on 40 cameras at the same time. However, I’m encountering an issue where I can’t achieve a minimum of 15 frames per second (fps) video output because Mediapipe is not utilizing the GPU for processing. I’m looking for ways to configure Mediapipe with Cuda accelerated computing in order to improve the output performance. Can you provide guidance on how to achieve this?
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
Processing camera samples from RTSP to Redis with ffmpeg | 1 | 1631 | May 25, 2022 | |
CUDA-OPENCV : low performances instead of high performances | 0 | 886 | April 13, 2016 | |
Using mediapipe gpu accelerated python solutions with cuda accelerated opencv | 6 | 6446 | December 8, 2021 | |
CUDA Toolkit for 1060 6gb ? | 0 | 711 | November 14, 2018 | |
Should I use GPU for generating thumbnails by FFmpeg? | 0 | 1411 | August 5, 2019 | |
Running Endoscopy Videos Need Help on running videos | 2 | 6080 | January 14, 2008 | |
real time 1080p video processing | 5 | 5478 | April 22, 2013 | |
Performance | 0 | 319 | June 15, 2019 | |
Software acceleration using GPU | 2 | 410 | May 25, 2019 | |
Efficient way to pipe capture video from webcam to cuda | 0 | 3103 | March 26, 2009 |