BetterPortrait

A lightweight macOS app that batch-processes portrait photos — detect faces, crop uniformly, remove backgrounds, and export. Built for when speakers submit inconsistent portrait photos and you need them all to look uniform.

Features

• Face Detection — Automatically detects faces using Apple's Vision framework
• Uniform Cropping — All photos are cropped so faces appear at a consistent size and position
• Background Removal — Person segmentation removes the original background and replaces it with a solid color of your choice
• Batch Processing — Import dozens of photos and process them all at once
• Multiple Aspect Ratios — Choose from 1:1, 3:4, 9:16, 4:3, or 16:9
• Two Sizing Modes
  • Best Fit — Faces are approximately similar size with better framing
  • Match Largest — All faces are exactly the same size, driven by the largest face
• No External Dependencies — Uses only Apple frameworks (Vision, Core Graphics, Core Image)

Requirements

• macOS 26 (Tahoe) or later
• Xcode 26 or later

Getting Started

1. Clone the repo
2. Open BetterPortrait/BetterPortrait.xcodeproj in Xcode
3. Build and run (Cmd+R)

Usage

1. Import — Drag and drop photos onto the main area, or click "Choose Files"
2. Configure — Use the left sidebar to select crop ratio, background color, and sizing mode
3. Process — Click the green "Process" button in the bottom-right panel
4. Export — Click "Export" and choose "Export All" or "Export Processed Only"

Photos where no face is detected are left untouched and marked with a warning.

Built With

• SwiftUI
• Vision framework (VNDetectFaceRectanglesRequest, VNGeneratePersonSegmentationRequest)
• Core Image (CIBlendWithMask for compositing)
• Core Graphics

Author

Haochen Zeng — @hzeng412

License

MIT