Hack to load custom depth map

hack.diff

diff --git a/modules/processing.py b/modules/processing.py
index 24c537d..0525cfd 100644
--- a/modules/processing.py
+++ b/modules/processing.py
@@ -156,20 +156,41 @@ class StableDiffusionProcessing():
         return image_conditioning
 
     def depth2img_image_conditioning(self, source_image):
-        # Use the AddMiDaS helper to Format our source image to suit the MiDaS model
-        transformer = AddMiDaS(model_type="dpt_hybrid")
-        transformed = transformer({"jpg": rearrange(source_image[0], "c h w -> h w c")})
-        midas_in = torch.from_numpy(transformed["midas_in"][None, ...]).to(device=shared.device)
-        midas_in = repeat(midas_in, "1 ... -> n ...", n=self.batch_size)
-
-        conditioning_image = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(source_image))
-        conditioning = torch.nn.functional.interpolate(
-            self.sd_model.depth_model(midas_in),
-            size=conditioning_image.shape[2:],
-            mode="bicubic",
-            align_corners=False,
-        )
-
+        conditioning = None
+
+        try:
+            script_dir = os.path.dirname(__file__)
+            rel_path = "../depthmap/current.png"
+            depth_img = Image.open(os.path.join(script_dir, rel_path))
+            depth_img = depth_img.convert("L")
+            depth_img = np.expand_dims(depth_img, axis=0)
+            depth_img = np.expand_dims(depth_img, axis=0).repeat(self.batch_size, axis=0)
+            depth_img = torch.from_numpy(depth_img)
+            depth_img = 2. * depth_img - 1.
+            depth_img = depth_img.to(shared.device)
+
+            conditioning_image = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(source_image))
+            conditioning = torch.nn.functional.interpolate(
+                depth_img,
+                size=conditioning_image.shape[2:],
+                mode="bicubic",
+                align_corners=False,
+            )
+        except:
+            # Use the AddMiDaS helper to Format our source image to suit the MiDaS model
+            transformer = AddMiDaS(model_type="dpt_hybrid")
+            transformed = transformer({"jpg": rearrange(source_image[0], "c h w -> h w c")})
+            midas_in = torch.from_numpy(transformed["midas_in"][None, ...]).to(device=shared.device)
+            midas_in = repeat(midas_in, "1 ... -> n ...", n=self.batch_size)
+
+            conditioning_image = self.sd_model.get_first_stage_encoding(self.sd_model.encode_first_stage(source_image))
+            conditioning = torch.nn.functional.interpolate(
+                self.sd_model.depth_model(midas_in),
+                size=conditioning_image.shape[2:],
+                mode="bicubic",
+                align_corners=False,
+            )
+       
         (depth_min, depth_max) = torch.aminmax(conditioning)
         conditioning = 2. * (conditioning - depth_min) / (depth_max - depth_min) - 1.
         return conditioning

Checks if there is a file at "../depthmap/current.png" and uses that otherwise falls back to MiDaS depth map.

Apply to webui using:

git apply hack.diff