Open-source alternative to Opus Clip, Vidyo.ai, Klap & SubMagic. Turn long-form YouTube videos into viral 9:16 shorts using LLM highlight detection, Whisper transcription, and auto vertical cropping — free, no watermarks, no per-clip credits.

[ICCV 2023] UniVTG: Towards Unified Video-Language Temporal Grounding

[EMNLP2024 Demo], [ICASSP 2025], [ICASSP 2026] A user-friendly library for reproducible video moment retrieval and highlight detection. It also supports audio moment retrieval.

Official pytorch repository for CG-DETR "Correlation-guided Query-Dependency Calibration in Video Representation Learning for Temporal Grounding"

FlashVTG: Feature Layering and Adaptive Score Handling Network for Video Temporal Grounding. (WACV2025)

AI-powered Auto-Clipper: automatically detects highlight segments from YouTube channels, Twitch/Kick live streams (via audience-retention data, chat spikes and timestamped comments), clips them with FFmpeg, generates subtitles using Whisper, and uploads them to a target YouTube channel. Includes a FastAPI + Celery backend, React + Tailwind frontend

本地 AI 高光剪辑工具：支持本地视频与主流视频链接导入，自动转写、打分、片段筛选与一键导出

A Codex Skill for tracking Video Moment Retrieval and Temporal Video Grounding research frontiers.

🎬 UMT: Unified Multi-modal Transformers for Joint Video Moment Retrieval and Highlight Detection (CVPR 2022)

🔥 长视频/直播回放一键切成爆款竖屏短视频:AI 找爆点、说话人分离、逐字字幕、竖屏重构、一键全托管,本地运行免费无水印 | Turn long videos & livestream replays into viral vertical shorts — AI highlight detection, speaker diarization, word-level captions, 9:16 auto-reframe. Free, local-first, no watermark.

Automatically turn Twitch VODs and YouTube videos into viral-ready clips using AI. Detects highlights via audio, chat, motion, and speech, then exports vertical videos with dynamic subtitles for TikTok, Reels, and Shorts.

Система для автоматического обнаружения основных моментов видеотрансляций / Automatic Highlights Detection of Broadcast Videos System

A machine learning system that detects highlight moments in Twitch livestreams by analyzing video, audio, and chat data, enabling efficient extraction of engaging content.