Closed AnnieXuY closed 4 years ago
The CPU count parameter affects the parallel processing of many files. The way the parsing of bitstream information works is that it can only run single-threaded, so any parallelism must be achieved by calculating the scores on many video files or segments in parallel. I hope this clarifies your question.
I understand and thank you for your response.
Hello guys, Thank you for providing this reference implementation. I did some test, and results of this project are indeed more accurate than vmaf. However time consuming is not an advantage. I found that the project provides the --cpu_count parameter, but no matter what number is set, the actual cpu_count is still 1. I wonder if this is a bug?
The test environment was as follows: Linux version 3.10.0-229.el7.x86_64 (gcc version 4.8.2 ) Python 3.6.6 cpu info: Intel(R) Xeon(R) CPU E5-2620 v2 @ 2.10GHz processor: 24 cpu cores : 6
The command line is as follows: poetry run p1204_3 --cpu_count 4 test_videos/test.mp4
How do I set it up to enable multiprocessing to improve the efficiency of the program?
Looking forward to a response.
BR, Annie