diff --git a/human/human_context.py b/human/human_context.py
index 0f1bc1f..193e0ea 100644
--- a/human/human_context.py
+++ b/human/human_context.py
@@ -9,7 +9,7 @@ from .audio_mal_handler import AudioMalHandler
 from .human_render import HumanRender
 from nlp import PunctuationSplit, DouBao
 from tts import TTSEdge, TTSAudioSplitHandle, TTSEdgeHttp
-from utils import load_avatar, get_device, object_stop, load_avatar_from_processed
+from utils import load_avatar, get_device, object_stop, load_avatar_from_processed, load_avatar_from_256_processed
 
 logger = logging.getLogger(__name__)
 current_file_path = os.path.dirname(os.path.abspath(__file__))
@@ -38,11 +38,16 @@ class HumanContext:
         base_path = os.path.join(current_file_path, '..')
         logger.info(f'base path:{base_path}')
         # full_images, face_frames, coord_frames = load_avatar(base_path, self._image_size, self._device)
-        full_images, face_frames, coord_frames = load_avatar_from_processed(base_path,
-                                                                            'wav2lip_avatar3')
+        # full_images, face_frames, coord_frames = load_avatar_from_processed(base_path,
+        #                                                                     'wav2lip_avatar3')
+        full_images, face_frames, coord_frames, align_frames, m_frames, inv_m_frames = load_avatar_from_256_processed(
+            base_path, 'wav2lip_avatar4', '26.pkl')
         self._frame_list_cycle = full_images
         self._face_list_cycle = face_frames
         self._coord_list_cycle = coord_frames
+        self._align_frames = align_frames
+        self._m_frames = m_frames
+        self._inv_m_frames = inv_m_frames
         face_images_length = len(self._face_list_cycle)
         logging.info(f'face images length: {face_images_length}')
         print(f'face images length: {face_images_length}')
diff --git a/utils/__init__.py b/utils/__init__.py
index 21e438c..2f3ca11 100644
--- a/utils/__init__.py
+++ b/utils/__init__.py
@@ -4,5 +4,5 @@ from .async_task_queue import AsyncTaskQueue
 from .sync_queue import SyncQueue
 from .utils import mirror_index, load_model, get_device, load_avatar, config_logging
 from .utils import read_image, object_stop
-from .utils import load_avatar_from_processed
+from .utils import load_avatar_from_processed, load_avatar_from_256_processed
 from .audio_utils import melspectrogram, save_wav
diff --git a/utils/utils.py b/utils/utils.py
index 174f74e..5d99a7a 100644
--- a/utils/utils.py
+++ b/utils/utils.py
@@ -210,6 +210,32 @@ def load_avatar_from_processed(base_path, avatar_name):
     return frame_list_cycle, face_list_cycle, coord_list_frames
 
 
+def load_avatar_from_256_processed(base_path, avatar_name, pkl):
+    avatar_path = os.path.join(base_path, 'data', 'avatars', avatar_name, pkl)
+    print(f'load avatar from processed:{avatar_path}')
+
+    with open(avatar_path, "rb") as f:
+        avatar_data = pickle.load(f)
+
+    face_list_cycle = []
+    frame_list_cycle = []
+    coord_list_frames = []
+    align_frames = []
+    m_frames = []
+    inv_m_frames = []
+
+    frame_info_list = avatar_data['frame_info_list']
+    for frame_info in frame_info_list:
+        face_list_cycle.append(frame_info['img'])
+        frame_list_cycle.append(frame_info['frame'])
+        coord_list_frames.append(frame_info['coords'])
+        align_frames.append(frame_info['align_frame'])
+        m_frames.append(frame_info['m'])
+        inv_m_frames.append(frame_info['inv_m'])
+
+    return frame_list_cycle, face_list_cycle, coord_list_frames, align_frames, m_frames, inv_m_frames
+
+
 def config_logging(file_name: str, console_level: int = logging.INFO, file_level: int = logging.DEBUG):
     file_handler = logging.FileHandler(file_name, mode='a', encoding="utf8")
     file_handler.setFormatter(logging.Formatter(