-
首先非常感谢作者的开源,但是我在使用tiny剪枝时遇到了错误,BrokenPipeError: [Errno 32] Broken pipe RuntimeError: The "freeze_support()" line can be omitted if the program大致是线程之类的错误,我用的使自己训练的四分类网络就是标准的yolo-tiny只更改了yolo层符合我的四个输出…
-
Hello everybody. I have been trying lately to detect small (~30 px size objects) in a 1920x1080 image. The another problem is that they are *way too sparse*, we have about one or two objects per image…
-
Hi @AlexeyAB,
I'm training yolov3 (tiny and standard) on my custom data set with my custom object classes. In a simple way it is detecting a dog or a cat in a frame of a video sequence.
I see poor m…
-
I am retraining yolov3 to detect custom class along with 20 existing classes using VOc 2007 and 2012 dataset. I followed instructions on
> "https://github.com/AlexeyAB/darknet#how-to-train-to-dete…
-
Hello @AlexeyAB ,
I try to run this command,
```
./darknet detector demo cfg/coco.data cfg/yolov3-tiny.cfg yolov3-tiny.weights gopro.MP4
```
Outputs are as follows.
```
CUDA-version: 1…
-
你好!
我想问下关于哪些卷积是可以裁剪的问题,在我看来并不是所有conv2d都可以裁剪,对于yolov3中每个shortcut所关联的2个卷积(-1,-3)应该是不能裁剪的吧?(因为shortcut是相加裁剪后层数不对应)
我这边计算年下yolov3所有带bn的卷积加起来一共有26304层,其中12544层和shortcut有关联,无法裁剪,将近一半。
由于你的repo中没有给出prune.…
-
I trained a yolov3 model (not the tiny version), and now I am trying to convert the saved yolo h5 file to Tensorflow pb, and so far I haven't had luck. This is how I am trying to do the conversion:
…
-
Can you fork some of the intermediate outputs of the backbone network into a few auxiliary layers which can be trained, using a cosine similarity loss function (or something similar), to output featur…
-
Using the latest darknet repo, trained my network using custom labels and images in this format:
`chars.data`:
```
classes=6
train=/mnt/c/Users/hello/Documents/github/darknet/chardata/train.txt
…
-
This is the plot i am getting
![chart - Copy](https://user-images.githubusercontent.com/33058092/71088410-5913d500-21c4-11ea-9833-d27fb92b27c8.png)
As you can see inspite of -map flag, only loss…