From 700c47a67492b1502265e5077c5be9ed70f8eb2a Mon Sep 17 00:00:00 2001 From: AUTOMATIC <16777216c@gmail.com> Date: Wed, 7 Sep 2022 17:00:51 +0300 Subject: big improvements to inpainting and outpainting --- modules/processing.py | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) (limited to 'modules/processing.py') diff --git a/modules/processing.py b/modules/processing.py index 49474b73..73b060f4 100644 --- a/modules/processing.py +++ b/modules/processing.py @@ -52,7 +52,7 @@ class StableDiffusionProcessing: self.overlay_images = overlay_images self.paste_to = None - def init(self): + def init(self, seed): pass def sample(self, x, conditioning, unconditional_conditioning): @@ -155,7 +155,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: precision_scope = torch.autocast if cmd_opts.precision == "autocast" else contextlib.nullcontext ema_scope = (contextlib.nullcontext if cmd_opts.lowvram else p.sd_model.ema_scope) with torch.no_grad(), precision_scope("cuda"), ema_scope(): - p.init() + p.init(seed=all_seeds[0]) if state.job_count == -1: state.job_count = p.n_iter @@ -240,7 +240,7 @@ def process_images(p: StableDiffusionProcessing) -> Processed: class StableDiffusionProcessingTxt2Img(StableDiffusionProcessing): sampler = None - def init(self): + def init(self, seed): self.sampler = samplers[self.sampler_index].constructor(self.sd_model) def sample(self, x, conditioning, unconditional_conditioning): @@ -320,7 +320,7 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing): self.mask = None self.nmask = None - def init(self): + def init(self, seed): self.sampler = samplers_for_img2img[self.sampler_index].constructor(self.sd_model) crop_region = None @@ -347,11 +347,13 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing): else: self.image_mask = images.resize_image(self.resize_mode, self.image_mask, self.width, self.height) np_mask = np.array(self.image_mask) - np_mask = 255 - np.clip((255 - np_mask.astype(np.float)) * 2, 0, 255).astype(np.uint8) + np_mask = np.clip((np_mask.astype(np.float)) * 2, 0, 255).astype(np.uint8) self.mask_for_overlay = Image.fromarray(np_mask) self.overlay_images = [] + latent_mask = self.latent_mask if self.latent_mask is not None else self.image_mask + imgs = [] for img in self.init_images: image = img.convert("RGB") @@ -361,7 +363,7 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing): if self.image_mask is not None: if self.inpainting_fill != 1: - image = fill(image, self.mask_for_overlay) + image = fill(image, latent_mask) image_masked = Image.new('RGBa', (image.width, image.height)) image_masked.paste(image.convert("RGBA").convert("RGBa"), mask=ImageOps.invert(self.mask_for_overlay.convert('L'))) @@ -394,17 +396,18 @@ class StableDiffusionProcessingImg2Img(StableDiffusionProcessing): self.init_latent = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(image)) if self.image_mask is not None: - init_mask = self.latent_mask if self.latent_mask is not None else self.image_mask + init_mask = latent_mask latmask = init_mask.convert('RGB').resize((self.init_latent.shape[3], self.init_latent.shape[2])) latmask = np.moveaxis(np.array(latmask, dtype=np.float64), 2, 0) / 255 latmask = latmask[0] + latmask = np.around(latmask) latmask = np.tile(latmask[None], (4, 1, 1)) self.mask = torch.asarray(1.0 - latmask).to(shared.device).type(self.sd_model.dtype) self.nmask = torch.asarray(latmask).to(shared.device).type(self.sd_model.dtype) if self.inpainting_fill == 2: - self.init_latent = self.init_latent * self.mask + create_random_tensors(self.init_latent.shape[1:], [self.seed + x + 1 for x in range(self.init_latent.shape[0])]) * self.nmask + self.init_latent = self.init_latent * self.mask + create_random_tensors(self.init_latent.shape[1:], [seed + x + 1 for x in range(self.init_latent.shape[0])]) * self.nmask elif self.inpainting_fill == 3: self.init_latent = self.init_latent * self.mask -- cgit v1.2.1