Enable metadata parsing for video uploads

Feature Request: Enable Metadata Parsing for Video Uploads

Summary:
Add support for automatic metadata parsing for video uploads on Citivai, similar to the current handling of image uploads.

Problem:
Currently, when users upload videos to Citivai, they must manually input metadata such as the original prompt, model, settings, and other relevant information. This process is time-consuming, error-prone, and inconsistent across users.

In contrast, image uploads automatically extract and parse embedded metadata (e.g., prompt, model, parameters), streamlining the user experience and improving data quality.

Proposed Solution:
Implement metadata parsing for supported video file formats (e.g., MP4, MOV) that may contain embedded metadata or associated sidecar files (e.g., JSON or XML). This will bring parity with image uploads and improve:

• Efficiency: Reduce manual input for users

• Accuracy: Prevent errors or omissions in metadata

• Consistency: Standardize metadata extraction across media types

Impact:
This feature would benefit all users uploading video content, especially creators working with AI-generated or prompt-driven video tools, and would align with current image handling behavior.

Optional Enhancements:

• Support parsing metadata from sidecar files if embedded metadata isn't available

• Provide a fallback UI for users to edit or confirm extracted metadata