yt_uniquifier

yt-uniquifier is a professional video processing tool built on ffmpeg designed to generate authorized derivatives of video content. It applies over 20 calibrated micro-transforms, including temporal frame jitter, pitch/tempo shifts, HDR-aware color adjustments, and parametric noise overlays, all composed into a single filter complex. Key capabilities include: - Interfaces: Provides a comprehensive CLI (`yt-uniq`), a WCAG 2.1 AA compliant PyQt6 desktop GUI, and a headless FastAPI web server with Docker support. - Processing Pipeline: Supports keyframe-aware and content-aware scene-cut splitting for resume-capable encoding, multi-vendor encoder detection (NVENC, QSV, AMF, VideoToolbox, x264, x265), and real HDR support via zscale. - Quality Assurance: Includes similarity diagnostics using pHash, audio Jaccard heuristics, and SSCD semantic-similarity models, generating detailed HTML QA reports with heatmaps. - Advanced Workflow: Features an experimental calibration tool for tuning profile intensity, a distributed batch system via shared filesystems (NFSv4), and a community profile marketplace for downloading validated presets. - Extensibility: Supports a third-party plugin system for registering new transforms and includes Whisper-driven soft subtitle injection.