configure model for pose detection

animation/animate-x/dwpose/wholebody.py

@@ -1,3 +1,4 @@
+import os
 import cv2
 import numpy as np
 
@@ -6,12 +7,15 @@ from dwpose.onnxdet import inference_detector
 from dwpose.onnxpose import inference_pose
 
 class Wholebody:
-    def __init__(self):
+    def __init__(self, model_directory=None):
         device = 'cuda' # 'cpu' #
         providers = ['CPUExecutionProvider'
                   ] if device == 'cpu' else ['CUDAExecutionProvider']
-        onnx_det =  'checkpoints/yolox_l.onnx' 
+        
-        onnx_pose = 'checkpoints/dw-ll_ucoco_384.onnx'
+        if not model_directory:
+            model_directory = 'checkpoints'
+        onnx_det =  os.path.join(model_directory, 'yolox_l.onnx')
+        onnx_pose = os.path.join(model_directory, 'dw-ll_ucoco_384.onnx')
 
         self.session_det = ort.InferenceSession(path_or_bytes=onnx_det, providers=providers)
         self.session_pose = ort.InferenceSession(path_or_bytes=onnx_pose, providers=providers)

animation/animate-x/process_data.py

@@ -89,9 +89,9 @@ def get_logger(name="essmc2"):
     return logger
 
 class DWposeDetector:
-    def __init__(self):
+    def __init__(self, model_directory=None):
 
-        self.pose_estimation = Wholebody()
+        self.pose_estimation = Wholebody(model_directory=model_directory)
 
     def __call__(self, oriImg):
         oriImg = oriImg.copy()
@@ -235,7 +235,7 @@ def mp_main(args):
     logger.info("There are {} videos for extracting poses".format(len(video_paths)))
 
     logger.info('LOAD: DW Pose Model')
-    dwpose_model = DWposeDetector()  
+    dwpose_model = DWposeDetector(model_directory=args.model_directory)  
 
     results_vis = []
     for i, file_path in enumerate(video_paths):
@@ -336,6 +336,7 @@ if __name__=='__main__':
         parser.add_argument("--saved_pose_dir", type=str, default="data/saved_pkl",)
         parser.add_argument("--saved_pose", type=str, default="data/saved_pose",)
         parser.add_argument("--saved_frame_dir", type=str, default="data/saved_frames",)
+        parser.add_argument("--model_directory", type=str, default="checkpoints",)
         args = parser.parse_args()
 
         return args

animation/animate-x/run_process_data.sh

@@ -6,5 +6,6 @@ python process_data.py \
     --source_video_paths data/videos/dance_1.mp4 \
     --saved_pose_dir data/saved_pkl \
     --saved_pose data/saved_pose \
-    --saved_frame_dir data/saved_frames
+    --saved_frame_dir data/saved_frames \
+    --model_directory checkpoints2