This dataset is divided into two subsets: real_data and synthetic_data.
The real_data subset is further divided into three splits {simple, moderate, challenging}, corresponding to three different videos. Each split contains the following:
A folder with the images of each frame of the video: imgs_{simple, moderate, challenging}
A folder containing a .txt file for each frame, where each row corresponds to an object of interest with its bounding box annotation in YOLO… See the full description on the dataset page:
https://huggingface.co/datasets/jjldo21/Laboratory.
