xiangbog commited on
Commit
86aa5b5
·
1 Parent(s): e320ab3

add metadata

Browse files
README.md CHANGED
@@ -1,3 +1,20 @@
1
- ---
2
- license: cc-by-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ ---
4
+
5
+
6
+ 1. Request access to the original DRAMA dataset at
7
+ https://usa.honda-ri.com/drama#Downloadthedataset
8
+
9
+ 2. Download the ZIP and extract the file
10
+ `integrated_output_v2.json`
11
+
12
+ 3. Ensure you have your existing
13
+ `drama_x_annotations.json` (with empty `image_path`/`video_path` fields) in the same folder.
14
+
15
+ 4. Run the population script:
16
+ ```bash
17
+ python populate_drama_x.py \
18
+ drama_x_annotated.json \
19
+ integrated_output_v2.json \
20
+ -o drama_x_annotations_populated.json
__pycache__/populate_drama_x.cpython-312.pyc ADDED
Binary file (2.65 kB). View file
 
drama_x_annotated.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
populate_drama_x.py ADDED
@@ -0,0 +1,53 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import json
2
+ import os
3
+ import argparse
4
+
5
+ def populate_paths(annotations_file: str, integrated_file: str, output_file: str):
6
+ # Load existing annotations
7
+ with open(annotations_file, 'r', encoding='utf-8') as f:
8
+ annotations = json.load(f)
9
+
10
+ # Load integrated_output_v2 entries
11
+ with open(integrated_file, 'r', encoding='utf-8') as f:
12
+ integrated = json.load(f)
13
+
14
+ # Update paths
15
+ for item in integrated:
16
+ img = item.get('s3_fileUrl', '')
17
+ vid = item.get('s3_instructionReference', '')
18
+ parts = img.rstrip('/').split('/')
19
+ if len(parts) < 2:
20
+ continue
21
+ clip_id = parts[-2] # e.g. 'clip_305_000786'
22
+ frame_file = parts[-1] # e.g. 'frame_000786.png'
23
+ frame_id = os.path.splitext(frame_file)[0]
24
+ key = f"{clip_id}_{frame_id}" # matches your JSON key
25
+
26
+ if key in annotations:
27
+ annotations[key]['image_path'] = img
28
+ annotations[key]['video_path'] = vid
29
+
30
+ # Write back
31
+ with open(output_file, 'w', encoding='utf-8') as f:
32
+ json.dump(annotations, f, indent=4, ensure_ascii=False)
33
+ print(f"Updated file written to {output_file}")
34
+
35
+ if __name__ == "__main__":
36
+ parser = argparse.ArgumentParser(
37
+ description="Populate image_path and video_path in drama_x_annotations.json"
38
+ )
39
+ parser.add_argument(
40
+ "annotations",
41
+ help="Path to existing drama_x_annotations.json"
42
+ )
43
+ parser.add_argument(
44
+ "integrated",
45
+ help="Path to integrated_output_v2.json"
46
+ )
47
+ parser.add_argument(
48
+ "-o", "--output",
49
+ default="drama_x_annotations_populated.json",
50
+ help="Where to save the updated annotations"
51
+ )
52
+ args = parser.parse_args()
53
+ populate_paths(args.annotations, args.integrated, args.output)