wutianze/dnndk-pynqz2

First lesson for you to use DNNDK, also it can be helpful for your AI learning

dnndk-pynqz2's Issues

Image Label on FPGA

Can you help me about Image Label on FPGA? I produced a .pb model file using Darknet tutorial. And then convert it to the .elf file. But while I convert , there is no information about input shape.(input shape = [?, ? ,?,?]). And I wrote in decent_q as ?,224,224,3. Then I uploaded .elf file into the ZCU102 -> adas detection -> model file. And I am modified necessary parameters classnum, anchor, biases etc. Then I run the program. But there is no label on the image or video or cam. Why? Could it be due to input shape?

關於將img燒進pynqz2後該如何使用

我已經將xilinx-pynqz2-dpu3.0-pynq2.5.img燒進pynqz2裡面

但是沒有看到下面這張圖裡的文件

我接下來應該把以下3個資料夾放進pynqz2裡面嗎?

另外問個問題
我在jupyter_notebooks/resnet50 裡面找到

執行到resnet50_cn.ipynb中的 run DPU and get the result

會造成
執行dpu後內核死亡所以自動重啟

我該如何使用這個dpu呢?

镜像文件无法启动

为什么我从百度云下载的PynqCar-DPU_v2.4_20191031.img文件，烧写到sd卡上板无法启动，板卡是pynq-z2。

安装包

您好？我想问一下您还有DNNDK的安装包吗？官网的好像下架了，在网上也没找到。

use yolo_prune to run video error

I ran the work in yolo_prune folder, and I got the result by ./yolo ./image/ i. But I ran into an error when I inference the model by the following command :
./yolo {VideoFileName.avi} v

The error message is :

Failed to query video capabilities: Inappropriate ioctl for device
libv4l2: error getting capabilities: Inappropriate ioctl for device
VIDEOIO ERROR: V4L: device ./video/road_trafifc.mp4: Unable to query number of channels
OpenCV Error: Unspecified error (GStreamer: your gstreamer installation is missing a required plugin
) in handlemessage, file /usr/src/debug/opencv/3.3+gitAUTOINC+87c27a074d_2a9d1b22ed_a62e20676a_34e4206aef_fccf7cd6a4-r0/git/modules/videoio/src/cap_gstreamer.cpp, line 1860
VIDEOIO(cvCreateCapture_GStreamer (CV_CAP_GSTREAMER_FILE, filename)): raised OpenCV exception:

/usr/src/debug/opencv/3.3+gitAUTOINC+87c27a074d_2a9d1b22ed_a62e20676a_34e4206aef_fccf7cd6a4-r0/git/modules/videoio/src/cap_gstreamer.cpp:1860: error: (-2) GStreamer: your gstreamer installation is missing a required plugin
 in function handlemessage

open video error!

It seems like missing some Gstreamer plugins, and I wonder how to solve it or tell me another way to run the video input successfully.
Thank you!

模型量化

您好！我想请教一下。我用yolov3得到一个230M的.h5文件，然后转成.pb也是230M，但是用dencent_q量化后，还是230M。按道理说，这个8bit量化后，模型不应该减小吗？我用这个量化后的230M的模型扔给DNNDK，报错了BACKEND][Check Failed: (shift_write >= -7) && (shift_write <= 7)][/tmp/DNNDK_Pipeline_dnnc/dnnc/submodules/asicv2comc/Operator/OperatorElew.cpp:63][DATA_OUTRANGE][Data value is out of range!]

The login problem

I can't log in with the username and password you provided.Could you please provide the correct one?please

有几个问题想请教一下

1. 我在赛灵思的Vitis AI Zoo里发现赛灵思优化了Tensorflow框架和Darknet框架下的yolo模型，目前想将训练好的模型移植到PYNQ-Z2板子上然后做目标检测，请问例如tf框架下的，赛灵思的Zoo里的模型和源作者yunyang1994开源的tf_yolov3差别是否会很大？

2. 能否将yunyang1994的tf_yolov3模型根据自己的需求训练好后，直接量化裁剪移植到PYNQ-Z2板子上呢？

3.如果2不可行的话，那么赛灵思是否提供了他们自己优化后的模型训练教程？

我刚刚开始入门，对这几个问题比较困惑，希望作者能解答一下，感激不尽！

dnndk下载地址似乎已经失效

参考build-host-dnndk.md文件中的dnndk_v3.0下载地址，但是链接已经失效无法下载，将我重定向到https://www.xilinx.com/support/download.html，但是我已经登录xilinx，我看到下载链接是
https://www.xilinx.com/member/forms/download/dnndk-eula-xef.html?filename=xlnx_dnndk_v3.0_190624.tar.gz
，请问是否还有其它方式能够下载，或者下载链接有误

DNNDK run error

####################################################
Warning:
The DPU in this TRD can only work 8 hours each time!
Please consult Sales for more details about this!
####################################################

[DNNDK] fail to open DPU device and exit ...

it's the error report,should I compile a new linux system ?
what should i do?please

制作镜像的文件从哪里下载呢？

前辈你好，如果要自己制作镜像的话，教程里的那些文件在哪里下载呢？

quant.sh

Hello, According to your instructions, I have successfully performed mnist....I'm running my own yolov3, but I have a problem quantifying pb files:ValueError: Input 0 of node pred_sbbox/strided_slice_8/aquant was passed int32 from pred_sbbox/strided_slice_8:0 incompatible with expected float.It hasn't been solved for a long time. Could you please help me.Thank you very much!

[DNNDK] fail to open DPU device and exit ...

请问在Running Applications on PYNQ-Z2的步骤下，2和3步骤具体如何执行呀？
这个rootfs.tar.gz在哪里呀？

1. Copy <PROJ ROOT>/pynq-z2-dpu/images/linux/image.ub, BOOT.BIN and rootfs.tar.gz to the sdcard directory.

cp ./images/linux/BOOT.BIN /media/$(whoami)/BOOT/
cp ./images/linux/image.ub /media/$(whoami)/BOOT/

2. Extract rootfs.tar.gz to rootfs ext4 partition

tar xzf ./images/linux/rootfs.tar.gz -C /media/$(whoami)/rootfs/

3. Copy yolo and dnndk to the sdcard and sync.

cp -r ../sdcard/* /media/$(whoami)/rootfs/home/root/
sync

Deploying a custom object detector

@wutianze I really appreciate your work, but I would like to get cleared of few details regarding my own network deployment on PYNQ Z2. I am completely new to FPGA and Xilinx software tools(But I understood the flow of deploying and using the DNNDK toolkit on a higher level). So my question is regarding the build-pynqz2-system. I have downloaded the image file from the link you have provided. if I have it, is it possible to skip all the steps mentioned in the build-pynqz2-system doc file? if not what should be done, I expect a clear explanation (how the pipeline flows ). I didn't mean your explanation is bad, I am not able to catch up.

Thanks in advance,
Zack

how to use dnnc?

if i have a new .elf file, and I download the pynq-z2-dpu140.zip from BaiDuWangPan. Can I use it directly? Should I do it all over again?

if i want run dnnc to generate a new .elf file, it need "xlnx_dnndk_v3.0_190624.zip" or "pynq-z2-dpu140.zip"？(your .tcl is vivado2018.3 I can't use it Mine is 2019.1)

freeze_graph

After converting .weight to h5, how does h5 freeze to pb?