Abdullah Ozcan

Alginment of MiDAS depth to ARCore real depth image problem

8

I am trying to align (convert relative depth units to real world depth) the MiDAS depth image to [ARCore ](https://codelabs.developers.google.com/codelabs/arcore-rawdepthapi#0) depth image. I am aware that there were already closed...

I get urls for such a search: `keyword= "car plate"` When I use it like `keyword= "car with plate"` nothing returns. Do you know what the problem is?

When I run the script "generate_components.py" I get this error below. Can you help? ``` Traceback (most recent call last): File "generate_components.py", line 112, in visualize_alpha(f'{sample_outdir}/depth_{component_index}_{component_name}.jpg', depths[component_index][0]) File "generate_components.py", line...

Hi there. When I run the demo, I get the warnings below. Like this one: `WARNING [09/23 05:51:06 d2.checkpoint.c2_model_loading]: Shape of roi_heads.box_predictor.bbox_pred.bias in checkpoint is torch.Size([4]), while shape of roi_heads.box_predictor.bbox_pred.bias...

Inpainting as object removal

1

First of all thanks for the great work. I have a question related to the inpainting with SD. If I want to remove an object completely from the scene, what...

Hi there. Thanks for this repository. I think there are some problems in model constructions, the model files and pretrained model dictionary seem to be not matching. I see problems...

Masked diffusion example

4

In the paper, there is the masked diffusion example (`Fig. 16`). I see that in the sampling functions the (eg. `ddim_sampling `function) mask image and the initial image (x0) are...

is the pretrained model for only faces?

1

Thanks for the work! Is the pretrained model is for only face pictures? If so, is there any other pretrained model for other objects, like box, bottle, shoe etc.?

When I use it with `stabilityai/stable-diffusion-2-inpainting` I get this error below. 

What to do to improve the quality of inference results

1

I have tried your trained model in some pictures. I have resized the pics to 512x1382 by keeping the aspect ratio. Also removed the background. Here are some results below....