videogui/process.py

import json
from collections import defaultdict

# Function to process the input JSON into desired structure
# Now groups by video_id with a single caption per video and list of frame entries

def process_ego4d_annotations(json_file_path, output_json_path=None):
    # Load the JSON data from file
    with open(json_file_path, 'r') as f:
        data = json.load(f)
    grouped = {}
    for video in data:
        video_id = video['video_id']
        caption = video['frames'][0]['caption']  # All frames have the same caption
        
        frames = []
        for frame in video['frames']:
            print(frame)
            frame_data = {
                "frame_path": frame['image_path'].replace('/home/shehan/workspace_grounding_lmm/Hanan/segment-anything-2/VideoGUI-High-Plan/video_gui_frames', 'videogui/frames/'),
                "mask_path": frame['mask'].replace('/home/shehan/workspace_grounding_lmm/Hanan/segment-anything-2/VideoGUI-High-Plan/', 'videogui/') \
                                if frame['mask'] is not None else frame['mask'],
                "points": [point[:2] for point in frame['clicked_points']]  # Remove the third element from clicked points
            }
            frames.append(frame_data)
        
        grouped[video_id] = {
            "caption": caption,
            "frames": frames
        }
        # Optionally save to output JSON file
    if output_json_path:
        with open(output_json_path, 'w') as outf:
            json.dump(grouped, outf, indent=2)

    return grouped


# Example usage
if __name__ == '__main__':
    input_json = '/share/data/drive_1/heakl/benchmark/annotated/videogui_data/VideoGUI_final.json'
    output_json = '/share/data/drive_1/heakl/benchmark/annotated/videogui_data/annotations.json'

    grouped_dict = process_ego4d_annotations(input_json, output_json)
    # Print nicely for verification
    import pprint
    pprint.pprint(grouped_dict)