Jo

@Dipet Annotation of partly non-visible objects, be it truncation or occlusion, is helpful when you want to train your model to "guess" the object size. Let's say you want to...

A workaround I am using for now is converting the 4 corners of my bounding box to "keypoints" -> transform them -> then create the transformed bounding boxes from the...

Is there any workaround for tf 2.2 for this? I am using conda to install tensorflow and currently only version 2.2 is available for linux.

I am still having that issue. Is there any progress on it? Did anyone maybe find some workaround for the time being?

I am having the same error when I save models during `on_train_begin()` or when I directly create it without training, thus only with the initialized weights. When training the model...

What fixes the above error for me is when I initialize the bias (in case the layer has `use_bias=True`) with anything else other than `"zero"`. ```python x = Conv2D(8, kernel_size=3)(x)...

I can only speak for tensorflow backend: There you will not be able to stop the debugger in the loss function while training. That is because how tensorflow is built....

Same issue with styled components extending rmwc components: ```javascript import * as React from 'react' import styled from 'styled-components' import { Button } from '@rmwc/button' const ButtonStyled = styled(Button)` `...

Follow these instructions https://www.tensorflow.org/install/install_windows. Install numpy and opencv. Download the weight files. Put them in the "weights" folder. Run any of the three python files.

For the face detection I have like 400ms. So about 2fps. In the paper they claim 40 fps and better. I really wonder what machine they used... EDIT: Okay, they...