JianJingGao

GTX3070，加载YOLOV5M模型，占用了3G左右的内存显存，有办法改善吗

请问怎样可以输入图片，输出人脸和对应label

8

请教您ZQCNN有方法吗，我看例程上是通过比对两张图相似度，，一定要输入一张图，去对比？ 期待您的解答，谢谢

ffmpeg not h264-nvmpi

1


### **multiple stream processing and multiple tracker are used，no trackek id for osd,please helpme should be how correctly configure ,thanks** `[source0] enable=1 type=4 #uri=file:///opt/nvidia/deepstream/deepstream/samples/streams/sample_1080p_h264.mp4 uri=rtsp://xxxxxxxxx num-sources=1 gpu-id=0 cudadec-memtype=0 latency =...

when i use libffmpeg awalys show [-s] [-r rate2] <file1.sw> <file2.sw>

1

when i use **libffmpeg_5.1.r106626_msvc15_x64** ,console awalys show [-s] [-r rate2] ,paragm exit when add a line include started go to wrong remove some ref such as libavcodec.lib , program normal...

vscmake编译main版本失败，麻烦大佬看看

1

环境是qt6，CMakeList如下 cmake_minimum_required(VERSION 3.20) project(FluentUI VERSION 0.1 LANGUAGES CXX) set(FLUENTUI_DIRECTORY ${CMAKE_CURRENT_LIST_DIR}) list(APPEND CMAKE_MODULE_PATH ${FLUENTUI_DIRECTORY}/.cmake/) include(GetGitRevisionDescription) option(FLUENTUI_BUILD_EXAMPLES "Build FluentUI demo applications." ON) option(FLUENTUI_BUILD_FRAMELESSHEPLER "Build FramelessHelper." ON) option(FLUENTUI_BUILD_STATIC_LIB "Build static library." OFF) set(Qt6_DIR...