Controlnet with Openvino SD Webui

[tuanakotta]

Hi there, I have managed to get the A1111 Webui running using the Arc A770 and the OpenVino acceleration script, one issue I am running into is the controlnet extension is completely ignored when I use the custom script, if I leave the script disabled controlnet is used but obviously image generation times are much slower, once I re-enable the script control net is completely ignored even though I can see the generated textures (canny image for example). Any help here would be greatly appreciated as can see in the script controlnet is referenced but I can't seem to find anything related to my issue online.

The controlnet info is even written into PNG info, but the output image is obviously not related to controlnet.

[Aznie_Intel]

Hi Tuanakotta,

Thanks for reaching out.

Which OpenVINO application are you running? You may check this Text-to-Image Generation with ControlNet Conditioning tutorial in Jupyter Notebook where canny edge detection is one of the target image for the sample.

Hope this helps.

Regards,

Aznie

[tuanakotta]

I was using openvino toolkit stable-diffusion-webui for automatic1111. Openvino version 2023.2.0. my interaction with AI image generation was limited to webui because I'm new to this.

[Aznie_Intel]

Hi Tuanakotta,

Are you using this stable-diffusion-webui repository? If you using the repo, I would suggest you post your query in the thread for better support.

On another note, stable diffusion demos are available in OpenVINO demos. You may check these 3 demos:

Stable Diffusion Text-to-Image Demo

Stable Diffusion v2.1 using Optimum-Intel OpenVINO

Stable Diffusion v2.1 using Optimum-Intel OpenVINO and multiple Intel Hardware

Regards,

Aznie

[Aznie_Intel]

Hi Tuanakotta,

This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.

Regards,

Aznie