fai-9/aa-as-is

aa-as-is's Issues

エラーメッセージ

はじめまして。githubに書き込むの初めてのため不備があったら申し訳ありません。

as isをi2iで使い、adetailerで以下のようにエラーが出ました。
ローカル版1.6使用です。

Traceback (most recent call last)
C:\Users\user\portable_stable\stable-diffusion-webui-1.6\extensions\adetailer-23.9.3\adetaile
r\traceback.py:139 in wrapper

136 width = width - 4 if width > 4 else None
137 console = Console(file=string, width=width)
138 try:
139return func(*args, **kwargs)
140 except Exception as e:
141tables = [
142 get_table(title, data)

C:\Users\user\portable_stable\stable-diffusion-webui-1.6\extensions\adetailer-23.9.3\scripts\
!adetailer.py:656 in postprocess_image

653for n, args in enumerate(arg_list):
654 if args.ad_model == "None":
655 continue
656 is_processed |= self._postprocess_image_inner(p, pp, args, n=n)
657
658 if is_processed:
659self.save_image(

C:\Users\user\portable_stable\stable-diffusion-webui-1.6\extensions\adetailer-23.9.3\scripts\
!adetailer.py:579 in _postprocess_image_inner

576kwargs["device"] = self.ultralytics_device
577
578 with change_torch_load():
579pred = predictor(ad_model, pp.image, args.ad_confidence, **kwargs)
580
581 masks = self.pred_preprocessing(pred, args)
582

C:\Users\user\portable_stable\stable-diffusion-webui-1.6\extensions\adetailer-23.9.3\adetaile
r\mediapipe.py:23 in mediapipe_predict

20    }  
21    if model_type in mapping:  
22  func = mapping[model_type]

23 return func(image, confidence)
24 msg = f"[-] ADetailer: Invalid mediapipe model type: {model_type}, Available: {list(
25 raise RuntimeError(msg)
26

C:\Users\user\portable_stable\stable-diffusion-webui-1.6\extensions\adetailer-23.9.3\adetaile
r\mediapipe.py:146 in mediapipe_face_mesh_eyes_only

143 static_image_mode=True, max_num_faces=20, min_detection_confidence=confidence
144 ) as face_mesh:
145 arr = np.array(image)
146 pred = face_mesh.process(arr)
147
148 if pred.multi_face_landmarks is None:
149return PredictOutput()

C:\Users\user\portable_stable\stable-diffusion-webui-1.6\venv\lib\site-packages\mediapipe\pyt
hon\solutions\face_mesh.py:125 in process

122 face landmarks on each detected face.
123 """
124
125 return super().process(input_data={'image': image})
126

C:\Users\user\portable_stable\stable-diffusion-webui-1.6\venv\lib\site-packages\mediapipe\pyt
hon\solution_base.py:354 in process

351 elif (input_stream_type == PacketDataType.IMAGE_FRAME or
352input_stream_type == PacketDataType.IMAGE):
353 if data.shape[2] != RGB_CHANNELS:
354 raise ValueError('Input image must contain three channel rgb data.')
355 self._graph.add_packet_to_input_stream(
356stream=stream_name,
357packet=self._make_packet(input_stream_type,
──
ValueError: Input image must contain three channel rgb data.

---

i2iに入れた画像はjpgでRGBです。
asisを使わない場合は問題なくadetailerが使えます。
私の環境のせいかもしれませんし、asisではなく
adetailerのほうに連絡したほうがいいかもしれませんが、
ご報告させていただきます。