Dataset Format

Our BANZ-FS is organized into several main directories, each containing various types of data essential for training, validating, and testing in multimedia and sign language analysis projects. The detailed organization facilitates straightforward access to data types across different experimental settings, as described below:

• Fingerspelling Recognition
  • Annotation
    • Train - Annotation file for training set (letter-level labels).
    • Val - Validation set annotations.
    • Test - Test set annotations.
  • Data
    • Fingerspelling_Recognition.zip - RGB videos for fingerspelling recognition.
• Fingerspelling Detection
  • Annotation
    • Train - Frame-level temporal boundaries of fingerspelling segments.
    • Val - Validation detection annotations.
    • Test - Test detection annotations.
  • Data
    • Fingerspelling_Detection.zip - Videos for fingerspelling detection task.
• Sign Language Translation
  • Annotation
    • Train - Parallel sign-to-text annotations for training.
    • Val - Validation set.
    • Test - Translation test annotations.
  • Data
    • Sign_Language_Translation.zip - Videos for sign language translation.