XtraHand

• 

• 

  landing_page

XtraHand is an AI-assisted pre-publish confidence layer for neurodivergent video creators and aspiring streamers. It addresses a barrier that is often overlooked: the anxiety between creating a video and finally publishing it.

For creators with ADHD, autism, dyslexia, or anxiety around uncertainty, the final review can become an exhausting cycle of rewatching, second-guessing, and worrying that a technical mistake was missed. XtraHand turns that ambiguous process into a clear, predictable precheck. It scans a video for issues such as clipped audio, long silences, black frames, frozen footage, visible private information, and unwanted borders. Instead of judging someone’s creativity, it highlights specific moments with timestamps and practical next steps.

The experience is designed to reduce cognitive load. Creators choose the checks that matter, adjust understandable sensitivity settings, and receive a short review queue rather than another full rewatch. Analysis happens locally, so unreleased footage remains private. The goal is not to make publishing perfect, it is to make pressing “Publish” feel safer.

XtraHand includes six filtering modes:

• Audio Clipping: Detects volume peaks that may cause distortion, such as shouting, impacts, or loud coughing.
• Long Silences: Identifies extended quiet sections that may have been left unedited.
• Black Frames: Finds black screens or video segments with missing visuals.
• Privacy Exposure: Flags possible appearances of desktops, windows, notifications, or private information for manual review.
• Frozen Picture: Detects images that remain frozen for too long.
• Black Borders: Identifies unwanted borders caused by formatting or export problems.

Each filter can be enabled separately, with adjustable sensitivity, minimum duration, and scanning frequency.

It removes friction between neurodivergent creators and creative expression. By providing reassurance before a video goes live, it can help more young people build confidence, experiment on camera, and see themselves as streamers rather than people who are “not ready yet.”

Built With

• ai
• ffmpeg
• numpy
• opencv
• python
• scipy
• streamlit