Source Diverse Sign Language Datasets

I am assembling a multilingual sign-language corpus for upcoming AI training and benchmarking. I need existing or newly collected material in ASL, ISL, BSL, or any other signed language. Both continuous signing sequences and isolated-sign clips are welcome, and I require that every file be paired with human-written English translations that already carry precise timestamps.

To move forward, please tell me exactly what you can supply: total hours or clip count, the number of unique signers, dialect or regional variants represented, camera setup, resolution, and your annotation format (e.g., JSON, ELAN, CSV). Clarify the consent process you followed and confirm that you can license the data for unrestricted commercial AI use.

If you are able to gather new recordings, outline your collection workflow, time frame, and per-hour or per-sign rate. For existing data, share a small watermarked sample along with your licensing terms and pricing model.

Acceptance Criteria
• Video files in a standard codec (MP4, MOV, or AVI)
• Aligned English transcripts with timestamps (SRT, CSV, or JSON)
• Documentation covering signer demographics, recording conditions, and annotation schema
• Signed licensing agreement granting perpetual commercial rights

I am ready to finalize quickly once I can review representative samples and confirm compliance.