dexter2406

I'm trying to convert the official *[monodepth2](https://github.com/nianticlabs/monodepth2)* model to Keras. The model is super simple and there's a `simple_test.py` which you could get a demo with one click. The weights...

Thanks for the work. I notice that it's by default *zero-padded*. I wonder if there's any way to achieve a different padding mode? For example, the `torch.nn.grid_sample()` supports "border" padding.

For example I want to use it on "person, car and bike", is it possible to train `deep cosine metrics` with multiple classes together for more "general" use?

小白求问个细节问题

1

在yolo_model.py 里，为啥有 ` grid_h, grid_w, num_boxes = map(int, out.shape[1: 4])`？ 按我的理解，out的尺寸应该是(13x13为例)：[13, 13, 3, 85]，如果要输出网格的长宽、bbox数量的话，那不应该是：`out.shape[0:3]` 吗？

The link `mono_640x192_pt(ours)` actually points to a file named with `...nopt`, and the other file is `...pt`. Is this just a wrong name or the file is wrong? And I...

No response in training process

4

Hi I found the program doesn't respond when I start training. The displayed information is like the following. There is no error report either. ``` np_resource = np.dtype([("resource", np.ubyte, 1)])...

ckpt file problem

2

Hi I wonder why your pretrained model doens't have "checkpoint number". Normally the ckpt for new-version tensorflow, the previous name should be changed to `model.ckpt-.data-00000-of-00001`. In an other model this...

请问不能用COCO格式训练吗？

1

首先谢谢您的工作！ 您提供了COCO上训练的 yolov4.weights，但只描述了VOC格式的训练方法，而且之前用的也是COCO，所以想请问能否用COCO格式训练？

FPS 和yolo3

1

(sorry 按错 还没打完就发送了) 我试了一下你的 Yolov3 和 v4 但速度都是大概 FPS=6 的样子 (GTX1050)，难道v4不应该更快一点吗

Where to get weights?

9

Hi, thanks for your work. But where could I got the weights file, e.g. `saved_model_0.19968511` ? I don't find link in your description. Thanks for your time!