Fix close-up selfie detection: pad margin so face-fills-frame photos are found (0 faces -> crude paste was the no-hair/skin bug)

Auto de-roll tilted source faces (eye-line alignment) before swap

Enable GPU for InsightFace: auto-add CUDA12 + cuDNN9 DLL dirs on Windows

Fix core swap: auto-download InsightFace buffalo_l + glasses in both paths

Fix all listed technical challenges across detection, landmarks, segmentation

fix: add explicit cv2.data import, add onnxruntime to CI deps

feat: auto-detect GPU/CPU for InsightFace and PyTorch (RTX 4060 locally, CPU on HF)

fix: correct retry logic for background model downloads, fix pyrightconfig extraPaths

fix: deduplicate InsightFace instance, add load-once sentinel to prevent retry loops

fix: use popup auth (redirect needs Firebase Hosting), fix BGR/RGB bug in InsightFace calls

Initial commit: DeepFace Studio - AI face swap web application