v0.1.6

• Added feature to pan a zoomed frame by holding Right Mouse button
• Improved Video Frame Seeking performance
• Fix for parameters getting lost when switching faces and workspaces not loading properly

Note: Python Requirements have also been updated to bump the version of PySide6. So, it is recommended to update your Python dependencies to reduce crashes some people were experiencing. (Especially Linux users)

v0.1.5

• Added batch selection of Input Faces by holding the Shift Key modifier
• Added option to Send the processed frames to a Virtual Camera (OBS Virtual Cam or Unity Capture) for the live streaming of the output in external applications
• Other bugfixes/improvements

v0.1.4

Added hotkeys

• F11: Toggle fullscreen
• V: Advance video by 1 frame
• C: Rewind video by 1 frame
• D: Advance video by 30 frames
• A: Rewind video by 30 frames
• Z: Seek to the start of the video
• Space: Play/Pause video
• R: Start/Stop recording
• F: Add/remove marker (Alt+F to remove)
• W: Move to the next marker
• Q: Move to the previous marker
• S: Toggle Swap faces

- Added context menu option to delete items from Target Media and Input Faces

- Added option to insert JPEG Compression artifacts into the output to add more realism when working with low quality target medias

- Update Parameters Handling to include Fallback Support

- Auto generation of thumbnails for all previously loaded target files

- Horizontal scrolling for Embeddings List

- Apply expression restorer before face restorer in the swapping pipeline

v0.1.3

• Added support for loading image files from paths with Unicode Characters (Non-English)
• Added buttons to load target/input media folders
• Last loaded path of Target, Input and Embeddings are now remembered when picking a different file/folder
• Parameters are now saved to the workspace file even if there was no target faces at the time of saving
• Other minor bugfixes/improvements

v0.1.2

• Fix invisible elements when using Dark-Blue theme
• Fix auto scroll issue on list widgets
• Address spacing issues in parameters panel
• Move Face Compare and Mask View toggles to the video panel
• 'Save Image' option now directly saves the image to the output folder
• Other minor bug fixes/improvements

v0.1.1

• Keep parameters and settings when changing target Image/Video

• Auto Swap feature added

• Tooltips improved for both dark & light themes ✨

• Keeps settings when switching media 🔄

• Updated labels and window titles 🏷️

• Improved UI and performance ⚡

• Default threads set to 2 to prevent memory issues 🚀

• Changed some defaults to address performance issues

• Minor fixes for video markers

• Fixed threading issues when using TensorRT-Engine provider.

• Fixed face restorer issues 🎭

• Fixed minor video slider issues 🎞️

• Lot of other minor and major bugfixes 🛠️

Now with a Windows Installer for users with Nvidia GPU