Merge branch 'dev' into main

.pre-commit-config.yaml

@@ -12,14 +12,13 @@ repos:
       - id: isort
         args: [--profile=black]
 
-  - repo: https://github.com/psf/black
-    rev: 23.3.0
-    hooks:
-      - id: black
-
   - repo: https://github.com/charliermarsh/ruff-pre-commit
-    # Ruff version.
     rev: "v0.0.265"
     hooks:
       - id: ruff
         args: [--fix, --exit-non-zero-on-fix]
+
+  - repo: https://github.com/psf/black
+    rev: 23.3.0
+    hooks:
+      - id: black

CHANGELOG.md

@@ -1,5 +1,12 @@
 # Changelog
 
+### 2023-05-09
+
+- v23.5.10
+- 선택한 스크립트만 ADetailer에 적용하는 옵션 추가, 기본값 `True`. 설정 탭에서 지정가능.
+  - 기본값: `dynamic_prompting,dynamic_thresholding,wildcards,wildcard_recursive`
+- `person_yolov8s-seg.pt` 모델 추가
+
 ### 2023-05-08
 
 - v23.5.9

README.md

@@ -36,6 +36,7 @@ On the ControlNet tab, select a ControlNet inpaint model and set the model weigh
 | mediapipe_face_short  | realistic face        | -                             | -                             |
 | hand_yolov8n.pt       | 2D / realistic hand   | 0.767                         | 0.505                         |
 | person_yolov8n-seg.pt | 2D / realistic person | 0.782 (bbox)<br/>0.761 (mask) | 0.555 (bbox)<br/>0.460 (mask) |
+| person_yolov8s-seg.pt | 2D / realistic person | 0.824 (bbox)<br/>0.809 (mask) | 0.605 (bbox)<br/>0.508 (mask) |
 
 The yolo models can be found on huggingface [Bingsu/adetailer](https://huggingface.co/Bingsu/adetailer).
 

adetailer/__version__.py

@@ -1 +1 @@
-__version__ = "23.5.9"
+__version__ = "23.5.10"

adetailer/common.py

@@ -39,6 +39,7 @@ def get_models(model_dir: Union[str, Path]) -> OrderedDict[str, Optional[str]]:
             "mediapipe_face_short": None,
             "hand_yolov8n.pt": hf_hub_download(repo_id, "hand_yolov8n.pt"),
             "person_yolov8n-seg.pt": hf_hub_download(repo_id, "person_yolov8n-seg.pt"),
+            "person_yolov8s-seg.pt": hf_hub_download(repo_id, "person_yolov8s-seg.pt"),
         }
     )
 
@@ -51,16 +52,16 @@ def get_models(model_dir: Union[str, Path]) -> OrderedDict[str, Optional[str]]:
 
 
 def create_mask_from_bbox(
-    image: Image.Image, bboxes: list[list[float]]
+    bboxes: list[list[float]], shape: tuple[int, int]
 ) -> list[Image.Image]:
     """
     Parameters
     ----------
-        image: Image.Image
-            The image to create the mask from
         bboxes: list[list[float]]
             list of [x1, y1, x2, y2]
             bounding boxes
+        shape: tuple[int, int]
+            shape of the image (width, height)
 
     Returns
     -------
@@ -70,7 +71,7 @@ def create_mask_from_bbox(
     """
     masks = []
     for bbox in bboxes:
-        mask = Image.new("L", image.size, 0)
+        mask = Image.new("L", shape, 0)
         mask_draw = ImageDraw.Draw(mask)
         mask_draw.rectangle(bbox, fill=255)
         masks.append(mask)

adetailer/mediapipe.py

@@ -45,7 +45,7 @@ def mediapipe_predict(
 
         bboxes.append([x1, y1, x2, y2])
 
-    masks = create_mask_from_bbox(image, bboxes)
+    masks = create_mask_from_bbox(bboxes, image.size)
     preview = Image.fromarray(preview_array)
 
     return PredictOutput(bboxes=bboxes, masks=masks, preview=preview)

adetailer/ultralytics.py

@@ -33,7 +33,7 @@ def ultralytics_predict(
     bboxes = bboxes.tolist()
 
     if pred[0].masks is None:
-        masks = create_mask_from_bbox(image, bboxes)
+        masks = create_mask_from_bbox(bboxes, image.size)
     else:
         masks = mask_to_pil(pred[0].masks.data, image.size)
     preview = pred[0].plot()
@@ -56,17 +56,17 @@ def ultralytics_check():
         print(message)
 
 
-def mask_to_pil(masks, orig_shape: tuple[int, int]) -> list[Image.Image]:
+def mask_to_pil(masks, shape: tuple[int, int]) -> list[Image.Image]:
     """
     Parameters
     ----------
     masks: torch.Tensor, dtype=torch.float32, shape=(N, H, W).
         The device can be CUDA, but `to_pil_image` takes care of that.
 
-    orig_shape: tuple[int, int]
+    shape: tuple[int, int]
         (width, height) of the original image
     """
     from torchvision.transforms.functional import to_pil_image
 
     n = masks.shape[0]
-    return [to_pil_image(masks[i], mode="L").resize(orig_shape) for i in range(n)]
+    return [to_pil_image(masks[i], mode="L").resize(shape) for i in range(n)]

scripts/!adetailer.py

@@ -448,6 +448,33 @@ class AfterDetailerScript(scripts.Script):
         params_txt = Path(data_path, "params.txt")
         params_txt.write_text(infotext, encoding="utf-8")
 
+    def script_filter(self, p, args: ADetailerArgs):
+        script_runner = copy(p.scripts)
+
+        ad_only_seleted_scripts = opts.data.get("ad_only_seleted_scripts", True)
+        if not ad_only_seleted_scripts:
+            return script_runner
+
+        default = "dynamic_prompting,dynamic_thresholding,wildcards,wildcard_recursive"
+        ad_script_names = opts.data.get("ad_script_names", default)
+        script_names_set = {
+            name
+            for script_name in ad_script_names.split(",")
+            for name in (script_name, script_name.strip())
+        }
+        if args.ad_controlnet_model != "None":
+            script_names_set.add("controlnet")
+
+        filtered_alwayson = []
+        for script_object in script_runner.alwayson_scripts:
+            filepath = script_object.filename
+            filename = Path(filepath).stem
+            if filename in script_names_set:
+                filtered_alwayson.append(script_object)
+
+        script_runner.alwayson_scripts = filtered_alwayson
+        return script_runner
+
     def get_i2i_p(self, p, args: ADetailerArgs, image):
         prompt, negative_prompt = self.get_prompt(p, args)
         seed, subseed = self.get_seed(p)
@@ -492,7 +519,7 @@ class AfterDetailerScript(scripts.Script):
             do_not_save_grid=True,
         )
 
-        i2i.scripts = copy(p.scripts)
+        i2i.scripts = self.script_filter(p, args)
         i2i.script_args = deepcopy(p.script_args)
         i2i._disable_adetailer = True
 
@@ -631,10 +658,10 @@ def on_ui_settings():
     shared.opts.add_option(
         "ad_max_models",
         shared.OptionInfo(
-            2,
-            "Max models",
-            gr.Slider,
-            {"minimum": 1, "maximum": 5, "step": 1},
+            default=2,
+            label="Max models",
+            component=gr.Slider,
+            component_args={"minimum": 1, "maximum": 5, "step": 1},
             section=section,
         ),
     )
@@ -649,5 +676,28 @@ def on_ui_settings():
         shared.OptionInfo(False, "Save images before ADetailer", section=section),
     )
 
+    shared.opts.add_option(
+        "ad_only_seleted_scripts",
+        shared.OptionInfo(
+            True, "Apply only selected scripts to ADetailer", section=section
+        ),
+    )
+
+    textbox_args = {
+        "placeholder": "comma-separated list of script names",
+        "interactive": True,
+    }
+
+    shared.opts.add_option(
+        "ad_script_names",
+        shared.OptionInfo(
+            default="dynamic_prompting,dynamic_thresholding,wildcards,wildcard_recursive",
+            label="Script names to apply to ADetailer (separated by comma)",
+            component=gr.Textbox,
+            component_args=textbox_args,
+            section=section,
+        ),
+    )
+
 
 script_callbacks.on_ui_settings(on_ui_settings)