fix annotations for SMPL-X_subset: remove frame-0 which is T-pose

SMPL-X_subset/Synbody_v1_0_sampled-smpl.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4f596de7243082720be6d0cceec616e566d2199382f7589b4a55c7f56b0deb77
-size 891643079
+oid sha256:5378e2b1a07cbc73ca6b8a682c20f31417880dcb2b60b0644dd6ab47c53348a4
+size 755794489

SMPL-X_subset/Synbody_v1_0_sampled-smplx.zip

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e02e372c35a5be9a4491177f570ffc9f4af40b82147ab01a6baeb1e9136e22d7
-size 2622572405
+oid sha256:3e40fbe308a4dc76de9333b438d62c40983fd3a5bd97121bca71e8e99e880964
+size 1941843369

SMPL-X_subset/visualize_2d.py

@@ -1,10 +1,10 @@
 """
 For examples:
 
->>> python release/visualize_2d.py \
-        --seq_dir synbody_v1_0/20230113/Downtown/LS_0114_004551_088/ \
-        --body_model_path {path_to_body_models} \
-        --save_path vis/LS_0114_004551_088.mp4
+>>> python visualize_2d.py \
+        --seq_dir synbody_v1_0/20230113/Downtown/LS_0114_004551_088_CAM002 \
+        --body_model {path_to_body_model} \
+        --save_path vis/LS_0114_004551_088_CAM002.mp4
 """
 
 from pathlib import Path
@@ -32,7 +32,7 @@ def load_data(seq_dir):
     seq_dir = Path(seq_dir)
     # load images
     frame_paths = sorted(seq_dir.glob('rgb/*.jpeg'))
-    images = [cv2.imread(p) for p in frame_paths]
+    images = [cv2.imread(str(p)) for p in frame_paths]
 
     # load parameters
     person_paths = sorted(seq_dir.glob('smplx/*.npz'))