diff options
Diffstat (limited to 'scripts')
| -rw-r--r-- | scripts/postprocessing_caption.py | 30 | ||||
| -rw-r--r-- | scripts/postprocessing_codeformer.py | 16 | ||||
| -rw-r--r-- | scripts/postprocessing_create_flipped_copies.py | 32 | ||||
| -rw-r--r-- | scripts/postprocessing_focal_crop.py | 54 | ||||
| -rw-r--r-- | scripts/postprocessing_gfpgan.py | 13 | ||||
| -rw-r--r-- | scripts/postprocessing_split_oversized.py | 71 | ||||
| -rw-r--r-- | scripts/postprocessing_upscale.py | 12 | ||||
| -rw-r--r-- | scripts/processing_autosized_crop.py | 64 | ||||
| -rw-r--r-- | scripts/soft_inpainting.py | 747 |
9 files changed, 1024 insertions, 15 deletions
diff --git a/scripts/postprocessing_caption.py b/scripts/postprocessing_caption.py new file mode 100644 index 00000000..243e3ad9 --- /dev/null +++ b/scripts/postprocessing_caption.py @@ -0,0 +1,30 @@ +from modules import scripts_postprocessing, ui_components, deepbooru, shared
+import gradio as gr
+
+
+class ScriptPostprocessingCeption(scripts_postprocessing.ScriptPostprocessing):
+ name = "Caption"
+ order = 4000
+
+ def ui(self):
+ with ui_components.InputAccordion(False, label="Caption") as enable:
+ option = gr.CheckboxGroup(value=["Deepbooru"], choices=["Deepbooru", "BLIP"], show_label=False)
+
+ return {
+ "enable": enable,
+ "option": option,
+ }
+
+ def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, option):
+ if not enable:
+ return
+
+ captions = [pp.caption]
+
+ if "Deepbooru" in option:
+ captions.append(deepbooru.model.tag(pp.image))
+
+ if "BLIP" in option:
+ captions.append(shared.interrogator.generate_caption(pp.image))
+
+ pp.caption = ", ".join([x for x in captions if x])
diff --git a/scripts/postprocessing_codeformer.py b/scripts/postprocessing_codeformer.py index a7d80d40..e1e156dd 100644 --- a/scripts/postprocessing_codeformer.py +++ b/scripts/postprocessing_codeformer.py @@ -1,28 +1,28 @@ from PIL import Image
import numpy as np
-from modules import scripts_postprocessing, codeformer_model
+from modules import scripts_postprocessing, codeformer_model, ui_components
import gradio as gr
-from modules.ui_components import FormRow
-
class ScriptPostprocessingCodeFormer(scripts_postprocessing.ScriptPostprocessing):
name = "CodeFormer"
order = 3000
def ui(self):
- with FormRow():
- codeformer_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="CodeFormer visibility", value=0, elem_id="extras_codeformer_visibility")
- codeformer_weight = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="CodeFormer weight (0 = maximum effect, 1 = minimum effect)", value=0, elem_id="extras_codeformer_weight")
+ with ui_components.InputAccordion(False, label="CodeFormer") as enable:
+ with gr.Row():
+ codeformer_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Visibility", value=1.0, elem_id="extras_codeformer_visibility")
+ codeformer_weight = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Weight (0 = maximum effect, 1 = minimum effect)", value=0, elem_id="extras_codeformer_weight")
return {
+ "enable": enable,
"codeformer_visibility": codeformer_visibility,
"codeformer_weight": codeformer_weight,
}
- def process(self, pp: scripts_postprocessing.PostprocessedImage, codeformer_visibility, codeformer_weight):
- if codeformer_visibility == 0:
+ def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, codeformer_visibility, codeformer_weight):
+ if codeformer_visibility == 0 or not enable:
return
restored_img = codeformer_model.codeformer.restore(np.array(pp.image, dtype=np.uint8), w=codeformer_weight)
diff --git a/scripts/postprocessing_create_flipped_copies.py b/scripts/postprocessing_create_flipped_copies.py new file mode 100644 index 00000000..3425571d --- /dev/null +++ b/scripts/postprocessing_create_flipped_copies.py @@ -0,0 +1,32 @@ +from PIL import ImageOps, Image
+
+from modules import scripts_postprocessing, ui_components
+import gradio as gr
+
+
+class ScriptPostprocessingCreateFlippedCopies(scripts_postprocessing.ScriptPostprocessing):
+ name = "Create flipped copies"
+ order = 4000
+
+ def ui(self):
+ with ui_components.InputAccordion(False, label="Create flipped copies") as enable:
+ with gr.Row():
+ option = gr.CheckboxGroup(value=["Horizontal"], choices=["Horizontal", "Vertical", "Both"], show_label=False)
+
+ return {
+ "enable": enable,
+ "option": option,
+ }
+
+ def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, option):
+ if not enable:
+ return
+
+ if "Horizontal" in option:
+ pp.extra_images.append(ImageOps.mirror(pp.image))
+
+ if "Vertical" in option:
+ pp.extra_images.append(pp.image.transpose(Image.Transpose.FLIP_TOP_BOTTOM))
+
+ if "Both" in option:
+ pp.extra_images.append(pp.image.transpose(Image.Transpose.FLIP_TOP_BOTTOM).transpose(Image.Transpose.FLIP_LEFT_RIGHT))
diff --git a/scripts/postprocessing_focal_crop.py b/scripts/postprocessing_focal_crop.py new file mode 100644 index 00000000..d3baf298 --- /dev/null +++ b/scripts/postprocessing_focal_crop.py @@ -0,0 +1,54 @@ +
+from modules import scripts_postprocessing, ui_components, errors
+import gradio as gr
+
+from modules.textual_inversion import autocrop
+
+
+class ScriptPostprocessingFocalCrop(scripts_postprocessing.ScriptPostprocessing):
+ name = "Auto focal point crop"
+ order = 4000
+
+ def ui(self):
+ with ui_components.InputAccordion(False, label="Auto focal point crop") as enable:
+ face_weight = gr.Slider(label='Focal point face weight', value=0.9, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_focal_crop_face_weight")
+ entropy_weight = gr.Slider(label='Focal point entropy weight', value=0.15, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_focal_crop_entropy_weight")
+ edges_weight = gr.Slider(label='Focal point edges weight', value=0.5, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_focal_crop_edges_weight")
+ debug = gr.Checkbox(label='Create debug image', elem_id="train_process_focal_crop_debug")
+
+ return {
+ "enable": enable,
+ "face_weight": face_weight,
+ "entropy_weight": entropy_weight,
+ "edges_weight": edges_weight,
+ "debug": debug,
+ }
+
+ def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, face_weight, entropy_weight, edges_weight, debug):
+ if not enable:
+ return
+
+ if not pp.shared.target_width or not pp.shared.target_height:
+ return
+
+ dnn_model_path = None
+ try:
+ dnn_model_path = autocrop.download_and_cache_models()
+ except Exception:
+ errors.report("Unable to load face detection model for auto crop selection. Falling back to lower quality haar method.", exc_info=True)
+
+ autocrop_settings = autocrop.Settings(
+ crop_width=pp.shared.target_width,
+ crop_height=pp.shared.target_height,
+ face_points_weight=face_weight,
+ entropy_points_weight=entropy_weight,
+ corner_points_weight=edges_weight,
+ annotate_image=debug,
+ dnn_model_path=dnn_model_path,
+ )
+
+ result, *others = autocrop.crop_image(pp.image, autocrop_settings)
+
+ pp.image = result
+ pp.extra_images = [pp.create_copy(x, nametags=["focal-crop-debug"], disable_processing=True) for x in others]
+
diff --git a/scripts/postprocessing_gfpgan.py b/scripts/postprocessing_gfpgan.py index d854f3f7..6e756605 100644 --- a/scripts/postprocessing_gfpgan.py +++ b/scripts/postprocessing_gfpgan.py @@ -1,26 +1,25 @@ from PIL import Image
import numpy as np
-from modules import scripts_postprocessing, gfpgan_model
+from modules import scripts_postprocessing, gfpgan_model, ui_components
import gradio as gr
-from modules.ui_components import FormRow
-
class ScriptPostprocessingGfpGan(scripts_postprocessing.ScriptPostprocessing):
name = "GFPGAN"
order = 2000
def ui(self):
- with FormRow():
- gfpgan_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="GFPGAN visibility", value=0, elem_id="extras_gfpgan_visibility")
+ with ui_components.InputAccordion(False, label="GFPGAN") as enable:
+ gfpgan_visibility = gr.Slider(minimum=0.0, maximum=1.0, step=0.001, label="Visibility", value=1.0, elem_id="extras_gfpgan_visibility")
return {
+ "enable": enable,
"gfpgan_visibility": gfpgan_visibility,
}
- def process(self, pp: scripts_postprocessing.PostprocessedImage, gfpgan_visibility):
- if gfpgan_visibility == 0:
+ def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, gfpgan_visibility):
+ if gfpgan_visibility == 0 or not enable:
return
restored_img = gfpgan_model.gfpgan_fix_faces(np.array(pp.image, dtype=np.uint8))
diff --git a/scripts/postprocessing_split_oversized.py b/scripts/postprocessing_split_oversized.py new file mode 100644 index 00000000..c4a03160 --- /dev/null +++ b/scripts/postprocessing_split_oversized.py @@ -0,0 +1,71 @@ +import math
+
+from modules import scripts_postprocessing, ui_components
+import gradio as gr
+
+
+def split_pic(image, inverse_xy, width, height, overlap_ratio):
+ if inverse_xy:
+ from_w, from_h = image.height, image.width
+ to_w, to_h = height, width
+ else:
+ from_w, from_h = image.width, image.height
+ to_w, to_h = width, height
+ h = from_h * to_w // from_w
+ if inverse_xy:
+ image = image.resize((h, to_w))
+ else:
+ image = image.resize((to_w, h))
+
+ split_count = math.ceil((h - to_h * overlap_ratio) / (to_h * (1.0 - overlap_ratio)))
+ y_step = (h - to_h) / (split_count - 1)
+ for i in range(split_count):
+ y = int(y_step * i)
+ if inverse_xy:
+ splitted = image.crop((y, 0, y + to_h, to_w))
+ else:
+ splitted = image.crop((0, y, to_w, y + to_h))
+ yield splitted
+
+
+class ScriptPostprocessingSplitOversized(scripts_postprocessing.ScriptPostprocessing):
+ name = "Split oversized images"
+ order = 4000
+
+ def ui(self):
+ with ui_components.InputAccordion(False, label="Split oversized images") as enable:
+ with gr.Row():
+ split_threshold = gr.Slider(label='Threshold', value=0.5, minimum=0.0, maximum=1.0, step=0.05, elem_id="postprocess_split_threshold")
+ overlap_ratio = gr.Slider(label='Overlap ratio', value=0.2, minimum=0.0, maximum=0.9, step=0.05, elem_id="postprocess_overlap_ratio")
+
+ return {
+ "enable": enable,
+ "split_threshold": split_threshold,
+ "overlap_ratio": overlap_ratio,
+ }
+
+ def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, split_threshold, overlap_ratio):
+ if not enable:
+ return
+
+ width = pp.shared.target_width
+ height = pp.shared.target_height
+
+ if not width or not height:
+ return
+
+ if pp.image.height > pp.image.width:
+ ratio = (pp.image.width * height) / (pp.image.height * width)
+ inverse_xy = False
+ else:
+ ratio = (pp.image.height * width) / (pp.image.width * height)
+ inverse_xy = True
+
+ if ratio >= 1.0 and ratio > split_threshold:
+ return
+
+ result, *others = split_pic(pp.image, inverse_xy, width, height, overlap_ratio)
+
+ pp.image = result
+ pp.extra_images = [pp.create_copy(x) for x in others]
+
diff --git a/scripts/postprocessing_upscale.py b/scripts/postprocessing_upscale.py index eb42a29e..ed709688 100644 --- a/scripts/postprocessing_upscale.py +++ b/scripts/postprocessing_upscale.py @@ -81,6 +81,14 @@ class ScriptPostprocessingUpscale(scripts_postprocessing.ScriptPostprocessing): return image
+ def process_firstpass(self, pp: scripts_postprocessing.PostprocessedImage, upscale_mode=1, upscale_by=2.0, upscale_to_width=None, upscale_to_height=None, upscale_crop=False, upscaler_1_name=None, upscaler_2_name=None, upscaler_2_visibility=0.0):
+ if upscale_mode == 1:
+ pp.shared.target_width = upscale_to_width
+ pp.shared.target_height = upscale_to_height
+ else:
+ pp.shared.target_width = int(pp.image.width * upscale_by)
+ pp.shared.target_height = int(pp.image.height * upscale_by)
+
def process(self, pp: scripts_postprocessing.PostprocessedImage, upscale_mode=1, upscale_by=2.0, upscale_to_width=None, upscale_to_height=None, upscale_crop=False, upscaler_1_name=None, upscaler_2_name=None, upscaler_2_visibility=0.0):
if upscaler_1_name == "None":
upscaler_1_name = None
@@ -126,6 +134,10 @@ class ScriptPostprocessingUpscaleSimple(ScriptPostprocessingUpscale): "upscaler_name": upscaler_name,
}
+ def process_firstpass(self, pp: scripts_postprocessing.PostprocessedImage, upscale_by=2.0, upscaler_name=None):
+ pp.shared.target_width = int(pp.image.width * upscale_by)
+ pp.shared.target_height = int(pp.image.height * upscale_by)
+
def process(self, pp: scripts_postprocessing.PostprocessedImage, upscale_by=2.0, upscaler_name=None):
if upscaler_name is None or upscaler_name == "None":
return
diff --git a/scripts/processing_autosized_crop.py b/scripts/processing_autosized_crop.py new file mode 100644 index 00000000..c0980226 --- /dev/null +++ b/scripts/processing_autosized_crop.py @@ -0,0 +1,64 @@ +from PIL import Image
+
+from modules import scripts_postprocessing, ui_components
+import gradio as gr
+
+
+def center_crop(image: Image, w: int, h: int):
+ iw, ih = image.size
+ if ih / h < iw / w:
+ sw = w * ih / h
+ box = (iw - sw) / 2, 0, iw - (iw - sw) / 2, ih
+ else:
+ sh = h * iw / w
+ box = 0, (ih - sh) / 2, iw, ih - (ih - sh) / 2
+ return image.resize((w, h), Image.Resampling.LANCZOS, box)
+
+
+def multicrop_pic(image: Image, mindim, maxdim, minarea, maxarea, objective, threshold):
+ iw, ih = image.size
+ err = lambda w, h: 1 - (lambda x: x if x < 1 else 1 / x)(iw / ih / (w / h))
+ wh = max(((w, h) for w in range(mindim, maxdim + 1, 64) for h in range(mindim, maxdim + 1, 64)
+ if minarea <= w * h <= maxarea and err(w, h) <= threshold),
+ key=lambda wh: (wh[0] * wh[1], -err(*wh))[::1 if objective == 'Maximize area' else -1],
+ default=None
+ )
+ return wh and center_crop(image, *wh)
+
+
+class ScriptPostprocessingAutosizedCrop(scripts_postprocessing.ScriptPostprocessing):
+ name = "Auto-sized crop"
+ order = 4000
+
+ def ui(self):
+ with ui_components.InputAccordion(False, label="Auto-sized crop") as enable:
+ gr.Markdown('Each image is center-cropped with an automatically chosen width and height.')
+ with gr.Row():
+ mindim = gr.Slider(minimum=64, maximum=2048, step=8, label="Dimension lower bound", value=384, elem_id="postprocess_multicrop_mindim")
+ maxdim = gr.Slider(minimum=64, maximum=2048, step=8, label="Dimension upper bound", value=768, elem_id="postprocess_multicrop_maxdim")
+ with gr.Row():
+ minarea = gr.Slider(minimum=64 * 64, maximum=2048 * 2048, step=1, label="Area lower bound", value=64 * 64, elem_id="postprocess_multicrop_minarea")
+ maxarea = gr.Slider(minimum=64 * 64, maximum=2048 * 2048, step=1, label="Area upper bound", value=640 * 640, elem_id="postprocess_multicrop_maxarea")
+ with gr.Row():
+ objective = gr.Radio(["Maximize area", "Minimize error"], value="Maximize area", label="Resizing objective", elem_id="postprocess_multicrop_objective")
+ threshold = gr.Slider(minimum=0, maximum=1, step=0.01, label="Error threshold", value=0.1, elem_id="postprocess_multicrop_threshold")
+
+ return {
+ "enable": enable,
+ "mindim": mindim,
+ "maxdim": maxdim,
+ "minarea": minarea,
+ "maxarea": maxarea,
+ "objective": objective,
+ "threshold": threshold,
+ }
+
+ def process(self, pp: scripts_postprocessing.PostprocessedImage, enable, mindim, maxdim, minarea, maxarea, objective, threshold):
+ if not enable:
+ return
+
+ cropped = multicrop_pic(pp.image, mindim, maxdim, minarea, maxarea, objective, threshold)
+ if cropped is not None:
+ pp.image = cropped
+ else:
+ print(f"skipped {pp.image.width}x{pp.image.height} image (can't find suitable size within error threshold)")
diff --git a/scripts/soft_inpainting.py b/scripts/soft_inpainting.py new file mode 100644 index 00000000..d9024344 --- /dev/null +++ b/scripts/soft_inpainting.py @@ -0,0 +1,747 @@ +import numpy as np +import gradio as gr +import math +from modules.ui_components import InputAccordion +import modules.scripts as scripts + + +class SoftInpaintingSettings: + def __init__(self, + mask_blend_power, + mask_blend_scale, + inpaint_detail_preservation, + composite_mask_influence, + composite_difference_threshold, + composite_difference_contrast): + self.mask_blend_power = mask_blend_power + self.mask_blend_scale = mask_blend_scale + self.inpaint_detail_preservation = inpaint_detail_preservation + self.composite_mask_influence = composite_mask_influence + self.composite_difference_threshold = composite_difference_threshold + self.composite_difference_contrast = composite_difference_contrast + + def add_generation_params(self, dest): + dest[enabled_gen_param_label] = True + dest[gen_param_labels.mask_blend_power] = self.mask_blend_power + dest[gen_param_labels.mask_blend_scale] = self.mask_blend_scale + dest[gen_param_labels.inpaint_detail_preservation] = self.inpaint_detail_preservation + dest[gen_param_labels.composite_mask_influence] = self.composite_mask_influence + dest[gen_param_labels.composite_difference_threshold] = self.composite_difference_threshold + dest[gen_param_labels.composite_difference_contrast] = self.composite_difference_contrast + + +# ------------------- Methods ------------------- + +def processing_uses_inpainting(p): + # TODO: Figure out a better way to determine if inpainting is being used by p + if getattr(p, "image_mask", None) is not None: + return True + + if getattr(p, "mask", None) is not None: + return True + + if getattr(p, "nmask", None) is not None: + return True + + return False + + +def latent_blend(settings, a, b, t): + """ + Interpolates two latent image representations according to the parameter t, + where the interpolated vectors' magnitudes are also interpolated separately. + The "detail_preservation" factor biases the magnitude interpolation towards + the larger of the two magnitudes. + """ + import torch + + # NOTE: We use inplace operations wherever possible. + + # [4][w][h] to [1][4][w][h] + t2 = t.unsqueeze(0) + # [4][w][h] to [1][1][w][h] - the [4] seem redundant. + t3 = t[0].unsqueeze(0).unsqueeze(0) + + one_minus_t2 = 1 - t2 + one_minus_t3 = 1 - t3 + + # Linearly interpolate the image vectors. + a_scaled = a * one_minus_t2 + b_scaled = b * t2 + image_interp = a_scaled + image_interp.add_(b_scaled) + result_type = image_interp.dtype + del a_scaled, b_scaled, t2, one_minus_t2 + + # Calculate the magnitude of the interpolated vectors. (We will remove this magnitude.) + # 64-bit operations are used here to allow large exponents. + current_magnitude = torch.norm(image_interp, p=2, dim=1, keepdim=True).to(torch.float64).add_(0.00001) + + # Interpolate the powered magnitudes, then un-power them (bring them back to a power of 1). + a_magnitude = torch.norm(a, p=2, dim=1, keepdim=True).to(torch.float64).pow_( + settings.inpaint_detail_preservation) * one_minus_t3 + b_magnitude = torch.norm(b, p=2, dim=1, keepdim=True).to(torch.float64).pow_( + settings.inpaint_detail_preservation) * t3 + desired_magnitude = a_magnitude + desired_magnitude.add_(b_magnitude).pow_(1 / settings.inpaint_detail_preservation) + del a_magnitude, b_magnitude, t3, one_minus_t3 + + # Change the linearly interpolated image vectors' magnitudes to the value we want. + # This is the last 64-bit operation. + image_interp_scaling_factor = desired_magnitude + image_interp_scaling_factor.div_(current_magnitude) + image_interp_scaling_factor = image_interp_scaling_factor.to(result_type) + image_interp_scaled = image_interp + image_interp_scaled.mul_(image_interp_scaling_factor) + del current_magnitude + del desired_magnitude + del image_interp + del image_interp_scaling_factor + del result_type + + return image_interp_scaled + + +def get_modified_nmask(settings, nmask, sigma): + """ + Converts a negative mask representing the transparency of the original latent vectors being overlayed + to a mask that is scaled according to the denoising strength for this step. + + Where: + 0 = fully opaque, infinite density, fully masked + 1 = fully transparent, zero density, fully unmasked + + We bring this transparency to a power, as this allows one to simulate N number of blending operations + where N can be any positive real value. Using this one can control the balance of influence between + the denoiser and the original latents according to the sigma value. + + NOTE: "mask" is not used + """ + import torch + return torch.pow(nmask, (sigma ** settings.mask_blend_power) * settings.mask_blend_scale) + + +def apply_adaptive_masks( + settings: SoftInpaintingSettings, + nmask, + latent_orig, + latent_processed, + overlay_images, + width, height, + paste_to): + import torch + import modules.processing as proc + import modules.images as images + from PIL import Image, ImageOps, ImageFilter + + # TODO: Bias the blending according to the latent mask, add adjustable parameter for bias control. + latent_mask = nmask[0].float() + # convert the original mask into a form we use to scale distances for thresholding + mask_scalar = 1 - (torch.clamp(latent_mask, min=0, max=1) ** (settings.mask_blend_scale / 2)) + mask_scalar = (0.5 * (1 - settings.composite_mask_influence) + + mask_scalar * settings.composite_mask_influence) + mask_scalar = mask_scalar / (1.00001 - mask_scalar) + mask_scalar = mask_scalar.cpu().numpy() + + latent_distance = torch.norm(latent_processed - latent_orig, p=2, dim=1) + + kernel, kernel_center = get_gaussian_kernel(stddev_radius=1.5, max_radius=2) + + masks_for_overlay = [] + + for i, (distance_map, overlay_image) in enumerate(zip(latent_distance, overlay_images)): + converted_mask = distance_map.float().cpu().numpy() + converted_mask = weighted_histogram_filter(converted_mask, kernel, kernel_center, + percentile_min=0.9, percentile_max=1, min_width=1) + converted_mask = weighted_histogram_filter(converted_mask, kernel, kernel_center, + percentile_min=0.25, percentile_max=0.75, min_width=1) + + # The distance at which opacity of original decreases to 50% + half_weighted_distance = settings.composite_difference_threshold * mask_scalar + converted_mask = converted_mask / half_weighted_distance + + converted_mask = 1 / (1 + converted_mask ** settings.composite_difference_contrast) + converted_mask = smootherstep(converted_mask) + converted_mask = 1 - converted_mask + converted_mask = 255. * converted_mask + converted_mask = converted_mask.astype(np.uint8) + converted_mask = Image.fromarray(converted_mask) + converted_mask = images.resize_image(2, converted_mask, width, height) + converted_mask = proc.create_binary_mask(converted_mask, round=False) + + # Remove aliasing artifacts using a gaussian blur. + converted_mask = converted_mask.filter(ImageFilter.GaussianBlur(radius=4)) + + # Expand the mask to fit the whole image if needed. + if paste_to is not None: + converted_mask = proc.uncrop(converted_mask, + (overlay_image.width, overlay_image.height), + paste_to) + + masks_for_overlay.append(converted_mask) + + image_masked = Image.new('RGBa', (overlay_image.width, overlay_image.height)) + image_masked.paste(overlay_image.convert("RGBA").convert("RGBa"), + mask=ImageOps.invert(converted_mask.convert('L'))) + + overlay_images[i] = image_masked.convert('RGBA') + + return masks_for_overlay + + +def apply_masks( + settings, + nmask, + overlay_images, + width, height, + paste_to): + import torch + import modules.processing as proc + import modules.images as images + from PIL import Image, ImageOps, ImageFilter + + converted_mask = nmask[0].float() + converted_mask = torch.clamp(converted_mask, min=0, max=1).pow_(settings.mask_blend_scale / 2) + converted_mask = 255. * converted_mask + converted_mask = converted_mask.cpu().numpy().astype(np.uint8) + converted_mask = Image.fromarray(converted_mask) + converted_mask = images.resize_image(2, converted_mask, width, height) + converted_mask = proc.create_binary_mask(converted_mask, round=False) + + # Remove aliasing artifacts using a gaussian blur. + converted_mask = converted_mask.filter(ImageFilter.GaussianBlur(radius=4)) + + # Expand the mask to fit the whole image if needed. + if paste_to is not None: + converted_mask = proc.uncrop(converted_mask, + (width, height), + paste_to) + + masks_for_overlay = [] + + for i, overlay_image in enumerate(overlay_images): + masks_for_overlay[i] = converted_mask + + image_masked = Image.new('RGBa', (overlay_image.width, overlay_image.height)) + image_masked.paste(overlay_image.convert("RGBA").convert("RGBa"), + mask=ImageOps.invert(converted_mask.convert('L'))) + + overlay_images[i] = image_masked.convert('RGBA') + + return masks_for_overlay + + +def weighted_histogram_filter(img, kernel, kernel_center, percentile_min=0.0, percentile_max=1.0, min_width=1.0): + """ + Generalization convolution filter capable of applying + weighted mean, median, maximum, and minimum filters + parametrically using an arbitrary kernel. + + Args: + img (nparray): + The image, a 2-D array of floats, to which the filter is being applied. + kernel (nparray): + The kernel, a 2-D array of floats. + kernel_center (nparray): + The kernel center coordinate, a 1-D array with two elements. + percentile_min (float): + The lower bound of the histogram window used by the filter, + from 0 to 1. + percentile_max (float): + The upper bound of the histogram window used by the filter, + from 0 to 1. + min_width (float): + The minimum size of the histogram window bounds, in weight units. + Must be greater than 0. + + Returns: + (nparray): A filtered copy of the input image "img", a 2-D array of floats. + """ + + # Converts an index tuple into a vector. + def vec(x): + return np.array(x) + + kernel_min = -kernel_center + kernel_max = vec(kernel.shape) - kernel_center + + def weighted_histogram_filter_single(idx): + idx = vec(idx) + min_index = np.maximum(0, idx + kernel_min) + max_index = np.minimum(vec(img.shape), idx + kernel_max) + window_shape = max_index - min_index + + class WeightedElement: + """ + An element of the histogram, its weight + and bounds. + """ + + def __init__(self, value, weight): + self.value: float = value + self.weight: float = weight + self.window_min: float = 0.0 + self.window_max: float = 1.0 + + # Collect the values in the image as WeightedElements, + # weighted by their corresponding kernel values. + values = [] + for window_tup in np.ndindex(tuple(window_shape)): + window_index = vec(window_tup) + image_index = window_index + min_index + centered_kernel_index = image_index - idx + kernel_index = centered_kernel_index + kernel_center + element = WeightedElement(img[tuple(image_index)], kernel[tuple(kernel_index)]) + values.append(element) + + def sort_key(x: WeightedElement): + return x.value + + values.sort(key=sort_key) + + # Calculate the height of the stack (sum) + # and each sample's range they occupy in the stack + sum = 0 + for i in range(len(values)): + values[i].window_min = sum + sum += values[i].weight + values[i].window_max = sum + + # Calculate what range of this stack ("window") + # we want to get the weighted average across. + window_min = sum * percentile_min + window_max = sum * percentile_max + window_width = window_max - window_min + + # Ensure the window is within the stack and at least a certain size. + if window_width < min_width: + window_center = (window_min + window_max) / 2 + window_min = window_center - min_width / 2 + window_max = window_center + min_width / 2 + + if window_max > sum: + window_max = sum + window_min = sum - min_width + + if window_min < 0: + window_min = 0 + window_max = min_width + + value = 0 + value_weight = 0 + + # Get the weighted average of all the samples + # that overlap with the window, weighted + # by the size of their overlap. + for i in range(len(values)): + if window_min >= values[i].window_max: + continue + if window_max <= values[i].window_min: + break + + s = max(window_min, values[i].window_min) + e = min(window_max, values[i].window_max) + w = e - s + + value += values[i].value * w + value_weight += w + + return value / value_weight if value_weight != 0 else 0 + + img_out = img.copy() + + # Apply the kernel operation over each pixel. + for index in np.ndindex(img.shape): + img_out[index] = weighted_histogram_filter_single(index) + + return img_out + + +def smoothstep(x): + """ + The smoothstep function, input should be clamped to 0-1 range. + Turns a diagonal line (f(x) = x) into a sigmoid-like curve. + """ + return x * x * (3 - 2 * x) + + +def smootherstep(x): + """ + The smootherstep function, input should be clamped to 0-1 range. + Turns a diagonal line (f(x) = x) into a sigmoid-like curve. + """ + return x * x * x * (x * (6 * x - 15) + 10) + + +def get_gaussian_kernel(stddev_radius=1.0, max_radius=2): + """ + Creates a Gaussian kernel with thresholded edges. + + Args: + stddev_radius (float): + Standard deviation of the gaussian kernel, in pixels. + max_radius (int): + The size of the filter kernel. The number of pixels is (max_radius*2+1) ** 2. + The kernel is thresholded so that any values one pixel beyond this radius + is weighted at 0. + + Returns: + (nparray, nparray): A kernel array (shape: (N, N)), its center coordinate (shape: (2)) + """ + + # Evaluates a 0-1 normalized gaussian function for a given square distance from the mean. + def gaussian(sqr_mag): + return math.exp(-sqr_mag / (stddev_radius * stddev_radius)) + + # Helper function for converting a tuple to an array. + def vec(x): + return np.array(x) + + """ + Since a gaussian is unbounded, we need to limit ourselves + to a finite range. + We taper the ends off at the end of that range so they equal zero + while preserving the maximum value of 1 at the mean. + """ + zero_radius = max_radius + 1.0 + gauss_zero = gaussian(zero_radius * zero_radius) + gauss_kernel_scale = 1 / (1 - gauss_zero) + + def gaussian_kernel_func(coordinate): + x = coordinate[0] ** 2.0 + coordinate[1] ** 2.0 + x = gaussian(x) + x -= gauss_zero + x *= gauss_kernel_scale + x = max(0.0, x) + return x + + size = max_radius * 2 + 1 + kernel_center = max_radius + kernel = np.zeros((size, size)) + + for index in np.ndindex(kernel.shape): + kernel[index] = gaussian_kernel_func(vec(index) - kernel_center) + + return kernel, kernel_center + + +# ------------------- Constants ------------------- + + +default = SoftInpaintingSettings(1, 0.5, 4, 0, 0.5, 2) + +enabled_ui_label = "Soft inpainting" +enabled_gen_param_label = "Soft inpainting enabled" +enabled_el_id = "soft_inpainting_enabled" + +ui_labels = SoftInpaintingSettings( + "Schedule bias", + "Preservation strength", + "Transition contrast boost", + "Mask influence", + "Difference threshold", + "Difference contrast") + +ui_info = SoftInpaintingSettings( + "Shifts when preservation of original content occurs during denoising.", + "How strongly partially masked content should be preserved.", + "Amplifies the contrast that may be lost in partially masked regions.", + "How strongly the original mask should bias the difference threshold.", + "How much an image region can change before the original pixels are not blended in anymore.", + "How sharp the transition should be between blended and not blended.") + +gen_param_labels = SoftInpaintingSettings( + "Soft inpainting schedule bias", + "Soft inpainting preservation strength", + "Soft inpainting transition contrast boost", + "Soft inpainting mask influence", + "Soft inpainting difference threshold", + "Soft inpainting difference contrast") + +el_ids = SoftInpaintingSettings( + "mask_blend_power", + "mask_blend_scale", + "inpaint_detail_preservation", + "composite_mask_influence", + "composite_difference_threshold", + "composite_difference_contrast") + + +# ------------------- Script ------------------- + + +class Script(scripts.Script): + def __init__(self): + self.section = "inpaint" + self.masks_for_overlay = None + self.overlay_images = None + + def title(self): + return "Soft Inpainting" + + def show(self, is_img2img): + return scripts.AlwaysVisible if is_img2img else False + + def ui(self, is_img2img): + if not is_img2img: + return + + with InputAccordion(False, label=enabled_ui_label, elem_id=enabled_el_id) as soft_inpainting_enabled: + with gr.Group(): + gr.Markdown( + """ + Soft inpainting allows you to **seamlessly blend original content with inpainted content** according to the mask opacity. + **High _Mask blur_** values are recommended! + """) + + power = \ + gr.Slider(label=ui_labels.mask_blend_power, + info=ui_info.mask_blend_power, + minimum=0, + maximum=8, + step=0.1, + value=default.mask_blend_power, + elem_id=el_ids.mask_blend_power) + scale = \ + gr.Slider(label=ui_labels.mask_blend_scale, + info=ui_info.mask_blend_scale, + minimum=0, + maximum=8, + step=0.05, + value=default.mask_blend_scale, + elem_id=el_ids.mask_blend_scale) + detail = \ + gr.Slider(label=ui_labels.inpaint_detail_preservation, + info=ui_info.inpaint_detail_preservation, + minimum=1, + maximum=32, + step=0.5, + value=default.inpaint_detail_preservation, + elem_id=el_ids.inpaint_detail_preservation) + + gr.Markdown( + """ + ### Pixel Composite Settings + """) + + mask_inf = \ + gr.Slider(label=ui_labels.composite_mask_influence, + info=ui_info.composite_mask_influence, + minimum=0, + maximum=1, + step=0.05, + value=default.composite_mask_influence, + elem_id=el_ids.composite_mask_influence) + + dif_thresh = \ + gr.Slider(label=ui_labels.composite_difference_threshold, + info=ui_info.composite_difference_threshold, + minimum=0, + maximum=8, + step=0.25, + value=default.composite_difference_threshold, + elem_id=el_ids.composite_difference_threshold) + + dif_contr = \ + gr.Slider(label=ui_labels.composite_difference_contrast, + info=ui_info.composite_difference_contrast, + minimum=0, + maximum=8, + step=0.25, + value=default.composite_difference_contrast, + elem_id=el_ids.composite_difference_contrast) + + with gr.Accordion("Help", open=False): + gr.Markdown( + f""" + ### {ui_labels.mask_blend_power} + + The blending strength of original content is scaled proportionally with the decreasing noise level values at each step (sigmas). + This ensures that the influence of the denoiser and original content preservation is roughly balanced at each step. + This balance can be shifted using this parameter, controlling whether earlier or later steps have stronger preservation. + + - **Below 1**: Stronger preservation near the end (with low sigma) + - **1**: Balanced (proportional to sigma) + - **Above 1**: Stronger preservation in the beginning (with high sigma) + """) + gr.Markdown( + f""" + ### {ui_labels.mask_blend_scale} + + Skews whether partially masked image regions should be more likely to preserve the original content or favor inpainted content. + This may need to be adjusted depending on the {ui_labels.mask_blend_power}, CFG Scale, prompt and Denoising strength. + + - **Low values**: Favors generated content. + - **High values**: Favors original content. + """) + gr.Markdown( + f""" + ### {ui_labels.inpaint_detail_preservation} + + This parameter controls how the original latent vectors and denoised latent vectors are interpolated. + With higher values, the magnitude of the resulting blended vector will be closer to the maximum of the two interpolated vectors. + This can prevent the loss of contrast that occurs with linear interpolation. + + - **Low values**: Softer blending, details may fade. + - **High values**: Stronger contrast, may over-saturate colors. + """) + + gr.Markdown( + """ + ## Pixel Composite Settings + + Masks are generated based on how much a part of the image changed after denoising. + These masks are used to blend the original and final images together. + If the difference is low, the original pixels are used instead of the pixels returned by the inpainting process. + """) + + gr.Markdown( + f""" + ### {ui_labels.composite_mask_influence} + + This parameter controls how much the mask should bias this sensitivity to difference. + + - **0**: Ignore the mask, only consider differences in image content. + - **1**: Follow the mask closely despite image content changes. + """) + + gr.Markdown( + f""" + ### {ui_labels.composite_difference_threshold} + + This value represents the difference at which the original pixels will have less than 50% opacity. + + - **Low values**: Two images patches must be almost the same in order to retain original pixels. + - **High values**: Two images patches can be very different and still retain original pixels. + """) + + gr.Markdown( + f""" + ### {ui_labels.composite_difference_contrast} + + This value represents the contrast between the opacity of the original and inpainted content. + + - **Low values**: The blend will be more gradual and have longer transitions, but may cause ghosting. + - **High values**: Ghosting will be less common, but transitions may be very sudden. + """) + + self.infotext_fields = [(soft_inpainting_enabled, enabled_gen_param_label), + (power, gen_param_labels.mask_blend_power), + (scale, gen_param_labels.mask_blend_scale), + (detail, gen_param_labels.inpaint_detail_preservation), + (mask_inf, gen_param_labels.composite_mask_influence), + (dif_thresh, gen_param_labels.composite_difference_threshold), + (dif_contr, gen_param_labels.composite_difference_contrast)] + + self.paste_field_names = [] + for _, field_name in self.infotext_fields: + self.paste_field_names.append(field_name) + + return [soft_inpainting_enabled, + power, + scale, + detail, + mask_inf, + dif_thresh, + dif_contr] + + def process(self, p, enabled, power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr): + if not enabled: + return + + if not processing_uses_inpainting(p): + return + + # Shut off the rounding it normally does. + p.mask_round = False + + settings = SoftInpaintingSettings(power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr) + + # p.extra_generation_params["Mask rounding"] = False + settings.add_generation_params(p.extra_generation_params) + + def on_mask_blend(self, p, mba: scripts.MaskBlendArgs, enabled, power, scale, detail_preservation, mask_inf, + dif_thresh, dif_contr): + if not enabled: + return + + if not processing_uses_inpainting(p): + return + + if mba.is_final_blend: + mba.blended_latent = mba.current_latent + return + + settings = SoftInpaintingSettings(power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr) + + # todo: Why is sigma 2D? Both values are the same. + mba.blended_latent = latent_blend(settings, + mba.init_latent, + mba.current_latent, + get_modified_nmask(settings, mba.nmask, mba.sigma[0])) + + def post_sample(self, p, ps: scripts.PostSampleArgs, enabled, power, scale, detail_preservation, mask_inf, + dif_thresh, dif_contr): + if not enabled: + return + + if not processing_uses_inpainting(p): + return + + nmask = getattr(p, "nmask", None) + if nmask is None: + return + + from modules import images + from modules.shared import opts + + settings = SoftInpaintingSettings(power, scale, detail_preservation, mask_inf, dif_thresh, dif_contr) + + # since the original code puts holes in the existing overlay images, + # we have to rebuild them. + self.overlay_images = [] + for img in p.init_images: + + image = images.flatten(img, opts.img2img_background_color) + + if p.paste_to is None and p.resize_mode != 3: + image = images.resize_image(p.resize_mode, image, p.width, p.height) + + self.overlay_images.append(image.convert('RGBA')) + + if len(p.init_images) == 1: + self.overlay_images = self.overlay_images * p.batch_size + + if getattr(ps.samples, 'already_decoded', False): + self.masks_for_overlay = apply_masks(settings=settings, + nmask=nmask, + overlay_images=self.overlay_images, + width=p.width, + height=p.height, + paste_to=p.paste_to) + else: + self.masks_for_overlay = apply_adaptive_masks(settings=settings, + nmask=nmask, + latent_orig=p.init_latent, + latent_processed=ps.samples, + overlay_images=self.overlay_images, + width=p.width, + height=p.height, + paste_to=p.paste_to) + + def postprocess_maskoverlay(self, p, ppmo: scripts.PostProcessMaskOverlayArgs, enabled, power, scale, + detail_preservation, mask_inf, dif_thresh, dif_contr): + if not enabled: + return + + if not processing_uses_inpainting(p): + return + + if self.masks_for_overlay is None: + return + + if self.overlay_images is None: + return + + ppmo.mask_for_overlay = self.masks_for_overlay[ppmo.index] + ppmo.overlay_image = self.overlay_images[ppmo.index] |