Merge pull request #9737 from AdjointOperator/master

add tiled inference support for ScuNET

extensions-builtin/ScuNET/scripts/scunet_model.py

@@ -5,11 +5,15 @@ import traceback
 import PIL.Image
 import numpy as np
 import torch
+from tqdm import tqdm
+
 from basicsr.utils.download_util import load_file_from_url
 
 import modules.upscaler
 from modules import devices, modelloader
 from scunet_model_arch import SCUNet as net
+from modules.shared import opts
+from modules import images
 
 
 class UpscalerScuNET(modules.upscaler.Upscaler):
@@ -42,28 +46,78 @@ class UpscalerScuNET(modules.upscaler.Upscaler):
             scalers.append(scaler_data2)
         self.scalers = scalers
 
-    def do_upscale(self, img: PIL.Image, selected_file):
+    @staticmethod
+    @torch.no_grad()
+    def tiled_inference(img, model):
+        # test the image tile by tile
+        h, w = img.shape[2:]
+        tile = opts.SCUNET_tile
+        tile_overlap = opts.SCUNET_tile_overlap
+        if tile == 0:
+            return model(img)
+
+        device = devices.get_device_for('scunet')
+        assert tile % 8 == 0, "tile size should be a multiple of window_size"
+        sf = 1
+
+        stride = tile - tile_overlap
+        h_idx_list = list(range(0, h - tile, stride)) + [h - tile]
+        w_idx_list = list(range(0, w - tile, stride)) + [w - tile]
+        E = torch.zeros(1, 3, h * sf, w * sf, dtype=img.dtype, device=device)
+        W = torch.zeros_like(E, dtype=devices.dtype, device=device)
+
+        with tqdm(total=len(h_idx_list) * len(w_idx_list), desc="ScuNET tiles") as pbar:
+            for h_idx in h_idx_list:
+
+                for w_idx in w_idx_list:
+
+                    in_patch = img[..., h_idx: h_idx + tile, w_idx: w_idx + tile]
+
+                    out_patch = model(in_patch)
+                    out_patch_mask = torch.ones_like(out_patch)
+
+                    E[
+                        ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf
+                    ].add_(out_patch)
+                    W[
+                        ..., h_idx * sf: (h_idx + tile) * sf, w_idx * sf: (w_idx + tile) * sf
+                    ].add_(out_patch_mask)
+                    pbar.update(1)
+        output = E.div_(W)
+
+        return output
+
+    def do_upscale(self, img: PIL.Image.Image, selected_file):
+
         torch.cuda.empty_cache()
 
         model = self.load_model(selected_file)
         if model is None:
+            print(f"ScuNET: Unable to load model from {selected_file}", file=sys.stderr)
             return img
 
         device = devices.get_device_for('scunet')
-        img = np.array(img)
-        img = img[:, :, ::-1]
-        img = np.moveaxis(img, 2, 0) / 255
-        img = torch.from_numpy(img).float()
-        img = img.unsqueeze(0).to(device)
+        tile = opts.SCUNET_tile
+        h, w = img.height, img.width
+        np_img = np.array(img)
+        np_img = np_img[:, :, ::-1]  # RGB to BGR
+        np_img = np_img.transpose((2, 0, 1)) / 255  # HWC to CHW
+        torch_img = torch.from_numpy(np_img).float().unsqueeze(0).to(device)  # type: ignore
 
-        with torch.no_grad():
-            output = model(img)
-        output = output.squeeze().float().cpu().clamp_(0, 1).numpy()
-        output = 255. * np.moveaxis(output, 0, 2)
-        output = output.astype(np.uint8)
-        output = output[:, :, ::-1]
+        if tile > h or tile > w:
+            _img = torch.zeros(1, 3, max(h, tile), max(w, tile), dtype=torch_img.dtype, device=torch_img.device)
+            _img[:, :, :h, :w] = torch_img # pad image
+            torch_img = _img
+
+        torch_output = self.tiled_inference(torch_img, model).squeeze(0)
+        torch_output = torch_output[:, :h * 1, :w * 1] # remove padding, if any
+        np_output: np.ndarray = torch_output.float().cpu().clamp_(0, 1).numpy()
+        del torch_img, torch_output
         torch.cuda.empty_cache()
-        return PIL.Image.fromarray(output, 'RGB')
+
+        output = np_output.transpose((1, 2, 0))  # CHW to HWC
+        output = output[:, :, ::-1]  # BGR to RGB
+        return PIL.Image.fromarray((output * 255).astype(np.uint8))
 
     def load_model(self, path: str):
         device = devices.get_device_for('scunet')
@@ -84,4 +138,3 @@ class UpscalerScuNET(modules.upscaler.Upscaler):
         model = model.to(device)
 
         return model
-

modules/shared.py

@@ -286,6 +286,8 @@ options_templates.update(options_section(('upscaling', "Upscaling"), {
     "ESRGAN_tile_overlap": OptionInfo(8, "Tile overlap, in pixels for ESRGAN upscalers. Low values = visible seam.", gr.Slider, {"minimum": 0, "maximum": 48, "step": 1}),
     "realesrgan_enabled_models": OptionInfo(["R-ESRGAN 4x+", "R-ESRGAN 4x+ Anime6B"], "Select which Real-ESRGAN models to show in the web UI. (Requires restart)", gr.CheckboxGroup, lambda: {"choices": shared_items.realesrgan_models_names()}),
     "upscaler_for_img2img": OptionInfo(None, "Upscaler for img2img", gr.Dropdown, lambda: {"choices": [x.name for x in sd_upscalers]}),
+    "SCUNET_tile": OptionInfo(256, "Tile size for SCUNET upscalers. 0 = no tiling.", gr.Slider, {"minimum": 0, "maximum": 512, "step": 16}),
+    "SCUNET_tile_overlap": OptionInfo(8, "Tile overlap, in pixels for SCUNET upscalers. Low values = visible seam.", gr.Slider, {"minimum": 0, "maximum": 64, "step": 1}),
 }))
 
 options_templates.update(options_section(('face-restoration', "Face restoration"), {