zishuo/ascend_community_projects
1
0
Fork 0
mirror of https://github.com/Ascend/ascend_community_projects.git synced 2026-04-22 23:07:09 +08:00
Code Issues Packages Projects Releases Wiki Activity

update MultiChannelVideoInfer/README.md.

Browse Source
This commit is contained in:
黄贺阳
2022-05-25 06:22:18 +00:00
committed by Gitee
parent 0c1a935f49
commit aeda1b2592
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
MultiChannelVideoInfer/README.md
+1 -1
View File
@@ -117,7 +117,7 @@ python3 main_infer.py
**测试帧率：**
本样例统计出每秒处理的帧数，输入为25帧率的视频，8路推理，3帧里面取一帧，理应每秒处理25 / 3 * 8 = 66.6帧，如统计每秒的帧率达到66.6则性能达标。本样例中提供4路推理和8路模型推理的pipeline，读者可以在main_infer.py中修改。经实验，更多路数模型推理可以提高并发，从而提高推理的性能。
本样例统计出每秒处理的帧数，输入为25帧率的视频，8路推理，3帧里面取一帧，理应每秒处理25 / 3 * 8 = 66.6帧，如统计每秒的帧率达到66.6则性能达标。本样例中提供4路推理和8路模型推理的pipeline，读者可以在main_infer.py中PIPELINE_PATH = b"./pipeline/multi_infer1_8.pipeline"中修改。经实验，更多路数模型推理可以提高并发，从而提高推理的性能。
注：输入视频帧率应高于25，否则无法发挥全部性能；性能测试可以把后面的print(retStr)代码注释掉，方便观察。