tuteming

Results 16 issues of tuteming

what is COMMON_INCLUDE in CMakeLists.txt line 34 ?

in AlexeyAB darknet, softmax_layer.c has two layers, softmax_layer and contrastive _layer. my problem is "how to use contrastive layer in classification?" ref in yolov4-tiny_contrastive.cfg and wiki CFG-Parameters-in-the-different-layers I should add...

請問一下如何將CMakeLists_yolov5-deepsort-tensorrt_win10.txt 改寫到QT5 .pro? 謝謝

我根據CMakeLists_yolov5-deepsort-tensorrt_win10.txt編譯完成後得到 deepsort.lib, yolosort.exe, yolosort.lib and yolov5_trt.lib 請問要如何使用他們 或進一步要如何得到yolov5s.engine和deepsort.engine,才能結合main.cpp的需求?

程式有2種編譯的方式:sln(test_dll.exe)及CMakeLists.txt(yolo-trt.exe)均在win10 vs2017 對於yolov4(custum weights)2種方式得到的答案都一樣,ok 我有一yolov4(custum config & weights)的精簡網路,依然以yolov4命名(config & weights) test_dll.exe & yolo-trt.exe 在先產生yolov4-kHALF-batch1.engine後,就會進行偵測,得到的答案都一樣,ok but,在已有engine時,程式會載入engine直接進行偵測, 1. test_dll.exe出現 Loading TRT Engine... Loading Complete! Assertion failed: get3DTensorVolume(m_Engine->getBindingDimensions(tensor.bindingIndex)) == tensor.volume && "Tensor volumes...

custom yolov4 主要在3個 [yolov]層定義classes數,在它的前一層set filters=3*(5+classes) 我想問的是yolov4的變型[Gaussian_yolo]層,在它的前一層set filters=3*(9+classes) 我知道tensorrt 沒有[Gaussian_yolo],我是否可以把[Gaussian_yolo]改名為[yolov] 然後在yolo.cpp及yolov4.cpp中把相關的5改成9. yolo.cpp line 290,692,845,859,1149,1245 yolov4.cpp line 30,32,34,36,39,46(?) 依然用 yolov4來產生engine.這樣行的通嗎? 如果你認為可以,上面註記的行數有沒有問題? 謝謝 煩請賜教

我使用sln在win10建置test_dll.exe,detector.dll成功,test_dll.exe也工作正常.yolov4-kFLOAT-batch1.eng也生成了. 因為sample_detector.cpp=>#include "class_timer.hpp" and #include "class_detector.h" 等於把整個extra and modules 全部都叫進去了 如果我要寫一個程式(已有yolov4-kFLOAT-batch1.eng),直接做偵測無須生成yolov4-kFLOAT-batch1.eng 請問我需要include modules 哪些檔案? 另外,我如果要用CMakeLists.txt在win10 compile, line 23 "stdc++fs" vs2017不支援,刪除它,compile ok, but run crash. how to solve this problem? 謝謝請賜教

yolov7&yolov7x have been released in https://github.com/AlexeyAB/darknet/issues/8595 can you update to yolov7 ? thanks.

where I can download the whole data set? include training, lable, and test data. Thanks