Change Log
- Readme Improvement: Thanks @ssawczyn
- auto Scan: Thanks @vick08
- Object Detection through rectangles: Any boxes without text such as icons.
- All the settings are moved to menu extras: Command+Control+Shift+S
- You can choose mode between GPT and OCR for both window and VOCursor scanning
- Scanning VOCursor will trigger OCR only if it classifies the image as a document
- Moved target window to settings menu.
- Moved save OCR result to the menu.
The GPT mode utilize GPT-4V, and they require your own OpenAI API key.
IMPORTANT: Location information from GPT-4V is extremely unreliable and inaccurate.
You can choose an image file from finder and open in VOCR to ask a question for GPT-4V.