Step 1: Download the ActivityNet v1.3 and THUMOS14 videos. For ActivityNet, you can submit a data request here. For THUMOS14, you can download it directly from the official website.
Step 2: Standardize all videos to MP4 format with a constant frame rate of 30fps using the script standardize_videos_to_constant_30fps_mp4.sh:
bash standardize_video_to_constant_30fps_mp4.sh <input_folder> <output_folder>
Step 3: Split the ActivityNet videos into three subfolders: train (10024 videos), valid (4926 videos), and test (5044 videos) using the official splits. Similarly, split THUMOS14 into valid (200 videos) and test (213 videos) subfolders.
Step 4: Generate metadata CSV files for each ActivityNet and THUMOS14 subset using the script generate_metadata_csv.py. This step is already pre-computed for the standardized ActivityNet and THUMOS14 videos and saved in the activitynet and thumos14 folders.
python generate_metadata_csv.py --video-folder <path/to/folder> --output-csv <output_filename.csv>