Skip to content

Commit ad2cd64

Browse files
committed
feat: v1.9.0 — Silero VAD, CrisperWhisper, RVM, emotion highlights, OTIO export
Silero VAD neural silence detection as alternative to energy threshold, with auto mode that tries AI first then falls back. CrisperWhisper verbatim filler word detection via HuggingFace transformers for far more accurate um/uh identification. Robust Video Matting backend for temporally consistent background removal. Florence-2 auto watermark region detection. Emotion-based highlight extraction via deepface facial expression analysis. Engagement scoring blends hook strength, emotional peak, pacing, and quotability with LLM scores. OpenTimelineIO export for universal NLE interchange (Resolve, Avid, FCP). Improved thumbnail scoring with composition balance and motion blur penalty. UXP panel gains project media browsing and AbortSignal polyfill. Frontend adds detection method selectors, scan debouncing, and capability cache reset.
1 parent ff84909 commit ad2cd64

38 files changed

Lines changed: 3651 additions & 408 deletions

CLAUDE.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -146,7 +146,7 @@
146146
- Lint: `ruff check opencut/` — codebase is fully clean, pre-commit enforces on every commit
147147

148148
## Version
149-
- Current: **v1.8.0**
149+
- Current: **v1.9.0**
150150
- All version strings: `pyproject.toml`, `__init__.py`, `CSXS/manifest.xml` (ExtensionBundleVersion + Version), `com.opencut.uxp/manifest.json`, `com.opencut.uxp/main.js` (VERSION const), `index.html` version display, README badge
151151
- Use `python scripts/sync_version.py --set X.Y.Z` to update all 18 targets at once (including UXP files)
152152

Install.ps1

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -155,7 +155,7 @@ Write-Host " \___/| .__/ \___|_| |_|\____\__,_|\__|" -ForegroundColor Cyan
155155
Write-Host " |_| " -ForegroundColor Cyan
156156
Write-Host ""
157157
Write-Host " Open Source Video Editing Automation" -ForegroundColor DarkGray
158-
Write-Host " Installer v1.8.0" -ForegroundColor DarkGray
158+
Write-Host " Installer v1.9.0" -ForegroundColor DarkGray
159159

160160
$isAdmin = Test-IsAdmin
161161
if ($isAdmin) {

OpenCut.iss

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@
22
; Fully self-contained installer — bundles server exe, ffmpeg, and CEP extension
33

44
#define MyAppName "OpenCut"
5-
#define MyAppVersion "1.8.0"
5+
#define MyAppVersion "1.9.0"
66
#define MyAppPublisher "SysAdminDoc"
77
#define MyAppURL "https://github.com/SysAdminDoc/OpenCut"
88

README.md

Lines changed: 12 additions & 9 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,11 @@
11
# OpenCut
22

3-
![Version](https://img.shields.io/badge/version-1.8.0-blue)
3+
![Version](https://img.shields.io/badge/version-1.9.0-blue)
44
![License](https://img.shields.io/badge/license-MIT-green)
55
![Platform](https://img.shields.io/badge/platform-Windows%2010%2F11-0078D4)
66
![Python](https://img.shields.io/badge/Python-3.9+-3776AB?logo=python&logoColor=white)
77
![Premiere Pro](https://img.shields.io/badge/Premiere%20Pro-2023+-9999FF?logo=adobepremierepro&logoColor=white)
8-
![Routes](https://img.shields.io/badge/API%20Routes-142-orange)
8+
![Routes](https://img.shields.io/badge/API%20Routes-148-orange)
99

1010
> A free, open-source Premiere Pro extension that brings AI-powered video editing automation, caption generation, audio processing, and visual effects -- all running locally on your machine. No subscriptions, no cloud, no API keys. Replaces the need for paid Premiere extensions.
1111
@@ -20,7 +20,7 @@
2020

2121
**Option A -- Installer exe (recommended):**
2222

23-
Download `OpenCut-Setup-1.8.0.exe` from [Releases](https://github.com/SysAdminDoc/OpenCut/releases) and run it. The installer handles everything: server exe, FFmpeg, CEP extension deployment, registry keys, desktop shortcut, and optional Whisper model download. **No Python or FFmpeg needed.**
23+
Download `OpenCut-Setup-1.9.0.exe` from [Releases](https://github.com/SysAdminDoc/OpenCut/releases) and run it. The installer handles everything: server exe, FFmpeg, CEP extension deployment, registry keys, desktop shortcut, and optional Whisper model download. **No Python or FFmpeg needed.**
2424

2525
**Option B -- From source (requires Python 3.9+ and [FFmpeg](https://ffmpeg.org/download.html) on PATH):**
2626

@@ -47,14 +47,14 @@ Clone the repo and run `Install.bat` as Administrator. It handles FFmpeg check,
4747

4848
## Features
4949

50-
OpenCut v1.8.0 includes **142+ API routes**, **8 panel tabs** with **50+ sub-tabs**, and covers every major video editing automation task — with workflow automation, project templates, i18n, Docker support, SQLite job persistence, 10 themes, keyboard shortcuts, and a UXP panel for Premiere Pro 25.6+.
50+
OpenCut v1.9.0 includes **148+ API routes**, **8 panel tabs** with **50+ sub-tabs**, and covers every major video editing automation task — with Silero VAD, CrisperWhisper filler detection, Robust Video Matting, emotion-based highlights, OpenTimelineIO export, Florence-2 watermark detection, and a UXP panel for Premiere Pro 25.6+.
5151

5252
### Cut & Clean
5353

5454
| Feature | Description | Engine |
5555
|---------|-------------|--------|
56-
| Silence Removal | Detect and remove silent segments with adjustable threshold | FFmpeg / VAD |
57-
| Filler Word Detection | Auto-detect and cut um, uh, like, you know, so, actually | WhisperX |
56+
| Silence Removal | Detect and remove silent segments. Three methods: energy threshold (classic), Silero VAD (neural AI), or auto (tries AI first) | FFmpeg / Silero VAD |
57+
| Filler Word Detection | Auto-detect and cut um, uh, like, you know, so, actually. Two backends: Whisper (fast) or CrisperWhisper (verbatim, most accurate) | WhisperX / CrisperWhisper |
5858
| Waveform Preview | Visual waveform with draggable threshold line synced to slider | FFmpeg PCM + Canvas |
5959
| Trim Tool | Set in/out points to extract a clip portion (stream copy or re-encode) | FFmpeg |
6060
| Full Pipeline | Combined silence + filler removal in one pass | Multi-stage |
@@ -115,21 +115,23 @@ OpenCut v1.8.0 includes **142+ API routes**, **8 panel tabs** with **50+ sub-tab
115115
| Auto-Crop Detect | Smart reframe anchor using cropdetect for talking-head content | FFmpeg cropdetect |
116116
| Merge / Concatenate | Join multiple clips (fast stream copy or re-encoded) | FFmpeg concat |
117117
| Side-by-Side Preview | Before/after frame comparison modal for effects | FFmpeg + base64 |
118-
| Watermark Removal | Remove logos via delogo or LaMA AI inpainting | FFmpeg / LaMA |
118+
| Watermark Removal | Remove logos via delogo or LaMA AI inpainting. Auto-detect watermark region with Florence-2 vision model | FFmpeg / LaMA / Florence-2 |
119119
| Color Match | Match the color profile of one clip to a reference clip using YCbCr histogram matching | OpenCV |
120120

121121
### AI & ML Tools
122122

123123
| Feature | Description | Engine |
124124
|---------|-------------|--------|
125125
| AI Upscaling | 3 tiers: Lanczos (fast), Real-ESRGAN (balanced), Video2x (premium) | FFmpeg / Real-ESRGAN |
126-
| Background Removal | Remove video backgrounds | rembg (U2-Net) |
126+
| Background Removal | Remove video backgrounds. Two backends: rembg (per-frame, more models) or Robust Video Matting (temporal consistency) | rembg / RVM |
127127
| Face Enhancement | Restore/upscale faces | GFPGAN |
128128
| Face Swap | Replace faces with reference image | InsightFace |
129129
| Style Transfer | Neural artistic style transfer | PyTorch models |
130-
| Auto Thumbnails | AI-scored frame extraction for thumbnails | OpenCV scoring |
130+
| Auto Thumbnails | AI-scored frame extraction with composition balance, center interest, and motion blur penalty | OpenCV scoring |
131131
| Auto Zoom Keyframes | Face-detected push-in zoom for talking-head content. Returns keyframes for the Premiere Pro Motion effect | OpenCV + ExtendScript |
132132
| Multicam Auto-Switching | Speaker diarization → camera cut list. Maps speakers to track indices and generates multicam edits | pyannote.audio + ExtendScript |
133+
| Emotion Highlights | Detect emotionally significant moments via facial expression analysis. Builds emotion curve to find peaks | deepface + OpenCV |
134+
| Engagement Scoring | Multi-signal highlight scoring: hook strength, emotional peak, pacing, quotability — blended with LLM scores | Built-in |
133135

134136
### Export & Batch
135137

@@ -140,6 +142,7 @@ OpenCut v1.8.0 includes **142+ API routes**, **8 panel tabs** with **50+ sub-tab
140142
| Batch Processing | Process multiple clips in parallel | ThreadPool |
141143
| Transcript Export | SRT, VTT, ASS, plain text, timestamped | Built-in |
142144
| Auto Thumbnails | AI-scored thumbnail candidates from video | OpenCV |
145+
| OpenTimelineIO Export | Universal timeline interchange for DaVinci Resolve, Avid, FCP, and any OTIO-compatible NLE | OpenTimelineIO |
143146

144147
### Panel UX
145148

0 commit comments

Comments
 (0)