How to use a script with Img2img API ?

[Expertaz]

Hello everyone!

I am new to AI art and a part of my thesis is about generating custom images.

I have attempted to use the Outpainting mk2 script within my Python code to outpaint an image, but I have not been successful. Is it even possible?

Here is my current code:

def option_img2img():
    option_payload = {
    "sd_model_checkpoint": "sd-v1-5-inpainting.ckpt",
    "do_not_add_watermark": True
    }

    requests.post(url=f'{url}/sdapi/v1/options', json=option_payload)

def img2img(p_env, n_p_mod, im):

    option_img2img()

    img_bytes = io.BytesIO()
    im.save(img_bytes, format='PNG')
    img_base64 = base64.b64encode(img_bytes.getvalue()).decode('utf-8')

    img2img_payload = {
        "init_images": [img_base64],
        "prompt": p_env,
        "negative_prompt": n_p_mod,
        "denoising_strength": 0.9,
        "width": 512,
        "height": 768,
        "cfg_scale": 7,
        "sampler_name": "Euler a",
        "restore_faces": False,
        "steps": 30,
        "script_args": ["outpainting mk2", ]
    }

    img2img_response = requests.post(url=f'{url}/sdapi/v1/img2img', json=img2img_payload)

    r = img2img_response.json()

    for i in r['images']:
        image = Image.open(io.BytesIO(base64.b64decode(i.split(",",1)[0])))

    return image

I'm not entirely certain if I added the 'script_args': ['outpainting mk2', ] correctly. Can you please check if this is the correct way to specify the Outpainting mk2 script?

Then the only thing remaining is to be able to specify the number of pixels to expand and the Outpainting direction, but I don't know how to do that. Can you please help me? Thank you very much!

[erer1243]

Set script_name to the name of the script. Set script_args to the arguments as they are defined in the script's run method.

[erer1243]

def run(self, p, _, pixels, mask_blur, direction, noise_q, color_variation):

from https://github.com/AUTOMATIC1111/stable-diffusion-webui/blob/master/scripts/outpainting_mk_2.py#L142

[erer1243]

(except for p, that is automatically the image you specified in init_images)

[erer1243]

eg

[Expertaz]

Hello ! Thank you a lot for your answer :)

[Erickrus]

I have developed another extensions to segment the image for inpaint, do you know how to write the send ? @Expertaz
The repo is: https://github.com/Erickrus/stable-diffusion-webui-segment-anything

[sunatte-saad]

Hi, could you help me send a payload for the inpaint img2img?
I want to send an image and its mask and then I want the prompt to generate graphics on the masked portion? I have been trying this for a while using API, but I was not able to<
Any leads?

here is the payload I have been trying, it does return an image but not as I want it to be

`
payload= {
"init_images": [cat],
"resize_mode": 0,
"denoising_strength": 0.75,

"mask": mask,
"mask_blur": 4,
"inpaint_full_res": 0,
"inpaint_full_res_padding": 32,
"inpainting_mask_invert": 1,
"inpainting_fill": 1,
"initial_noise_multiplier": 0,
"prompt": "mountain",
"seed": -1,
"batch_size": 1,
"steps": 20,
"cfg_scale": 7,
"width": 512,
"height": 512,
"restore_faces": False,
"do_not_save_samples": False,
"override_settings_restore_afterwards": True,
"include_init_images": True,
"send_images": True,
"save_images": False
}`