cmdbug / tnn_demo Goto Github PK
View Code? Open in Web Editor NEW🍉 移动端TNN部署学习笔记,支持Android与iOS。
License: GNU General Public License v3.0
🍉 移动端TNN部署学习笔记,支持Android与iOS。
License: GNU General Public License v3.0
使用的模型:https://github.com/ultralytics/yolov5/releases/download/v5.0/yolov5s.pt
转换onnx命令:python models/export.py --weights yolov5s.pt --img 640 --batch 1
转换TNN命令:python converter.py onnx2tnn yolov5s.onnx -o . -in 1,3,448,640
转换后的模型:https://www.lanzoui.com/iSktTqkfote
运行demo时报错:
我在“ https://github.com/DefTruth/lite.ai.toolkit ”这找到很多转好的tnn可用模型,如上图所示,但这些模型均不能正常运行,有的是闪退有的是fps特别低,请问一下大神更换模型都需要修改什么,谢谢!
你好我 根据安卓demo的cpp代码,只扣出调用tnn,测试图片的代码,在本机运行成功。
我自己训练一个模型,只修改为一类,别的默认官方的,请问为何提示无法调用。模型输出不对吗?
input_cvt_param.scale = {1.0 / 255, 1.0 / 255, 1.0 / 255, 0.0};
input_cvt_param.bias = {0.0, 0.0, 0.0, 0.0};
why scale is 1/255 ,and the last is 0.0?
bias is ?
您好, 我用官方给的yolov5s.pt转为tnn,测试结果正常。 但把自己训练的pt转为tnn后,精度下降很多(跟pytorch版本对比),这是哪里的原因呢?
请问必须会java吗? 请问怎样把自己的cmake过的算法 塞进去替换原有的呢?
核心的修改代码是那几个呢?
Hi. I am running your demo and feel confused about your YOLOv5 demo.
There is a "layers" parameter in the Yolov5.h:
TNN_Demo/iOS_TNN_Demo/TNNDemo/Yolov5.h
Line 87 in f361ba0
I wonder what's the meanings of these values?
你好,我使用华为手机,尝试使用GPU运行(CPU测试正常),但一直出错:Get OpenCL command queue failed! 请问大神是怎么跑通GPU的
我想问一下,TNN人脸检测配准-腾讯优图实现这个例子,有关于人脸特征分布图吗,获取到的key_points 是随机的还是按固定规则排序的
nandodet官网直接export出onnx,然后直接转成的tnn可以直接用在此处?
# yolov5.h文件
std::vector<std::string> labels{"prohibitory", "warning", "mandatory"};
private:
std::shared_ptr<tnn::TNN> net;
std::shared_ptr<TNN_NS::Instance> instance;
TNN_NS::DeviceType device_type;
int net_width = 640;
int net_height = 640;
int num_class = 3;
prohibitory
,结果显示的是coco中的person
,求大神解答一下Hello. I trained a nanodet model from
https://github.com/RangiLyu/nanodet,
and an error occurred to me when processing
auto status = NanoDet::instance->GetCommandQueue(&command_queue);
:
2021-10-03 22:27:42.800142+0800 TNNDemo[7059:1059735] new NanoDet
E/tnn: virtual tnn::Status tnn::DefaultNetwork::InitLayers(tnn::NetStructure *, tnn::NetResource *) [File /Users/TURUI/Desktop/TNN/source/tnn/core/default_network.cc][Line 236] Error Init layer Conv_0 (err: 4096 or 0x1000)
2021-10-03 22:27:42.839470+0800 TNNDemo[7059:1059735] GPU initialization failed, switch to CPU
E/tnn: virtual tnn::Status tnn::DefaultNetwork::InitLayers(tnn::NetStructure *, tnn::NetResource *) [File /Users/TURUI/Desktop/TNN/source/tnn/core/default_network.cc][Line 236] Error Init layer Conv_0 (err: 4096 or 0x1000)
2021-10-03 22:27:42.845144+0800 TNNDemo[7059:1059735] TNN init failed 4096
(lldb)
It seems that the error is caused by initialization. Do you have any ideas to fix it?
Hello thanks for sharing these samples - can you also add the other model cards of landmarks and hair segmentation?
使用您的权重文件没有任何问题,当使用自己的自己的权重文件(通过https://convertmodel.com转换,出现了一下错误。
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.