Commit Graph

549 Commits

Author SHA1 Message Date
Kenneth Estanislao b6b6c741a2 Revert "Merge pull request #1710 from ozp3/amd-dml-optimization"
This reverts commit 1b240a45fd, reversing
changes made to d9a5500bdf.
2026-04-01 22:33:01 +08:00
Kenneth Estanislao 1b240a45fd Merge pull request #1710 from ozp3/amd-dml-optimization
AMD GPU (DirectML) Optimization for Live Mode
2026-04-01 22:29:43 +08:00
ozp3 ecf02d0640 Delete DeepLiveCam.lnk
remove lnk and bat files as requested
2026-04-01 16:46:28 +03:00
ozp3 0cbc9f126f Delete run-dml.bat
remove lnk and bat files as requested
2026-04-01 16:45:31 +03:00
Kenneth Estanislao d9a5500bdf Merge pull request #1713 from TeachDian/fix-1705-wsl-onnxruntime-gpu 2026-03-29 04:54:34 +08:00
TeachDian 86134b6e1d Fix #1705: Update onnxruntime-gpu requirement to 1.23.2 for WSL compatibility 2026-03-29 04:46:48 +08:00
ozp3 fbd1cc5973 docs: add AMD DML optimization notes to README 2026-03-28 13:16:43 +03:00
ozp3 eac2ad2307 feat: AMD DML optimization - GPU face detection, detection throttle, pre-load fix 2026-03-28 13:09:20 +03:00
Kenneth Estanislao 9e6f30c0a4 silenced deprecation 2026-03-27 21:35:27 +08:00
Kenneth Estanislao 97321a740d Update face_analyser.py
320 was over optimized, put back to 640
2026-03-27 21:24:19 +08:00
Kenneth Estanislao f5f7ac7764 Revise README for clarity and formatting
Updated README to remove emoji and clarify GPU support details.
2026-03-23 10:02:50 +08:00
Kenneth Estanislao 77d3492eef Add download link for models in README
Added a section for downloading models from Hugging Face.
2026-03-13 23:39:46 +08:00
Kenneth Estanislao 8e3d6e7c65 Add emoji to project title in README
Just want to add an emoji 😝
2026-03-13 22:17:32 +08:00
Kenneth Estanislao ee9699ee70 Happy 80k!
2.1 Released!

- Face randomizer added!
2026-03-13 22:09:18 +08:00
Kenneth Estanislao 3c8b259a3f Some edits on the UI
- Grouped the face enhancers
- Make the mouth mask just a slider
- Removed the redundant switches
2026-03-13 22:03:28 +08:00
Kenneth Estanislao 30b27c2b71 Update Quick Start section to v2.7 beta 2026-03-12 02:40:52 +08:00
Kenneth Estanislao 0d8f3b1f82 Fix on vulnerability report
https://github.com/hacksider/Deep-Live-Cam/issues/1695
2.7-beta
2026-03-06 23:26:48 +08:00
KRSHH 6e9e7addf2 Update press section with recent media mentions 2026-03-03 21:16:56 +05:30
Kenneth Estanislao 0c7e871bfc Merge pull request #1689 from laurigates/pr/base-ui-tooltips
feat(ui): add hover tooltips to all controls
2026-02-28 02:41:07 +08:00
Lauri Gates e340b0da8a feat(ui): add hover tooltips to all controls
Add ToolTip class (modules/ui_tooltip.py) and wire descriptive hover
tooltips onto every button, switch, slider, and dropdown in the main
window. Tooltips appear after a 500ms hover delay and are clamped to
screen bounds.

This requires no new dependencies — ToolTip uses only customtkinter.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-24 21:41:24 +02:00
Kenneth Estanislao d0f81ed755 Merge pull request #1671 from laurigates/pr/fix-macos-camera-enum
fix(macos): replace cv2_enumerate_cameras with safe bounded loop
2026-02-24 14:29:00 +08:00
Kenneth Estanislao de01b28802 Merge pull request #1678 from laurigates/pr/perf-opacity-handling
perf(face-swapper): optimize opacity handling and frame copies
2026-02-24 14:28:17 +08:00
Lauri Gates b645d5e60b fix(macos): replace cv2_enumerate_cameras with safe bounded loop
cv2_enumerate_cameras(CAP_AVFOUNDATION) probes indices 0-99 through
OpenCV's AVFoundation backend, which intermittently segfaults (exit
code 139) when invalid device indices are probed. Replace with a
bounded cv2.VideoCapture loop (range(10)) that safely skips
unavailable indices.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-23 17:22:35 +02:00
Kenneth Estanislao 31b3a97003 Merge pull request #1680 from laurigates/pr/perf-float32-buffer-reuse
perf(processing): optimize post-processing with float32 and buffer reuse
2026-02-23 15:13:03 +08:00
Kenneth Estanislao e3b46e83b7 Merge pull request #1669 from laurigates/pr/feat-gpen-enhancers
feat: add GPEN-BFR 256 and 512 ONNX face enhancers
2026-02-23 15:05:44 +08:00
Lauri Gates e93fb95903 perf(processing): optimize post-processing with float32 and buffer reuse
- Replace float64 with float32 in apply_mouth_area() blending masks —
  float32 provides sufficient precision for 8-bit image blending and
  halves memory bandwidth
- Use float32 in apply_mask_area() mask computations
- Vectorize hull padding loop in create_face_mask() (face_masking.py)
  replacing per-point Python loop with NumPy array operations
- Fix apply_color_transfer() to use proper [0,1] LAB conversion —
  cv2.cvtColor with float32 input expects [0,1] range, not [0,255]
- Pre-compute inverse masks to avoid repeated (1.0 - mask) subtraction
- Use np.broadcast_to instead of np.repeat for face mask expansion

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-22 21:27:31 +02:00
Lauri Gates aabf41050a perf(face-swapper): optimize opacity handling and frame copies
Move opacity calculation before frame copy to skip the copy when
opacity is 1.0 (common case). Add early return path for full opacity.
Clear PREVIOUS_FRAME_RESULT instead of caching when interpolation
is disabled.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-22 21:12:02 +02:00
Lauri Gates e57116de68 feat: add GPEN-BFR 256 and 512 ONNX face enhancers
Add two new face enhancement processors using GPEN-BFR ONNX models
at 256x256 and 512x512 resolutions. Models auto-download on first
use from GitHub releases. Integrates into existing frame processor
pipeline alongside GFPGAN enhancer with UI toggle switches.

- modules/paths.py: Shared path constants module
- modules/processors/frame/_onnx_enhancer.py: ONNX enhancement utilities
- modules/processors/frame/face_enhancer_gpen256.py: GPEN-BFR 256 processor
- modules/processors/frame/face_enhancer_gpen512.py: GPEN-BFR 512 processor
- modules/core.py: Add GPEN choices to --frame-processor CLI arg
- modules/globals.py: Add GPEN entries to fp_ui toggle dict
- modules/ui.py: Add GPEN toggle switches and processing integration

Closes #1663

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-22 19:39:12 +02:00
Kenneth Estanislao d5338a3eae Update version in README and add contributor 2026-02-23 01:02:22 +08:00
Kenneth Estanislao 7ec3a4be29 Merge pull request #1665 from laurigates/pr/perf-pipeline-threading
perf(ui): decouple face detection from swap in live webcam pipeline
2026-02-23 00:59:22 +08:00
Lauri Gates ca6cba9311 perf(ui): decouple face detection from swap in live webcam pipeline
Add a dedicated detection thread that runs face detection continuously
on the latest captured frame and publishes results to a shared dict.
The processing/swap thread reads cached detection results instead of
running detection inline, so it never blocks on the 15-30ms detection
cost.

Architecture change: 2 threads → 3 threads
  Before: capture → [detect + swap] → display
  After:  capture → swap (uses cached detections) → display
                  ↘ detect (async, writes to shared cache) ↗

Also replaces the blocking while/ROOT.update() display loop with
ROOT.after()-based scheduling, which avoids Tk event loop re-entrancy
issues and UI freezes.

Closes #1664
2026-02-22 18:41:47 +02:00
Kenneth Estanislao d89385457e Merge pull request #1659 from laurigates/pr/fix-tk9-compat
fix(ui): patch CTkOptionMenu for Tk 9.0 compatibility
2026-02-23 00:13:47 +08:00
Kenneth Estanislao b015f0099f Update GFPGANv1.4 download link to ONNX format 2026-02-23 00:03:37 +08:00
Kenneth Estanislao e56a79222e Merge branch 'main' of https://github.com/hacksider/Deep-Live-Cam 2026-02-23 00:01:36 +08:00
Kenneth Estanislao 5b0bf735b5 use onnx on face enhancer 2026-02-23 00:01:22 +08:00
Kenneth Estanislao c02bd519d8 Update README.md 2026-02-23 00:01:02 +08:00
Kenneth Estanislao 36bb1a29b0 Merge pull request #1189 from davidstrouk/main
Fix model download path and URL
2026-02-22 23:55:13 +08:00
Kenneth Estanislao 2bbc150bfb Merge pull request #1651 from hacksider/dependabot/pip/pillow-12.1.1
Bump pillow from 11.1.0 to 12.1.1
2026-02-22 18:01:34 +08:00
Lauri Gates a1722c7b2e fix(ui): patch CTkOptionMenu for Tk 9.0 compatibility
In Tk 9.0, Menu.index("end") returns "" instead of raising TclError
on empty menus. CustomTkinter's DropdownMenu._add_menu_commands
doesn't handle this case, causing a crash when creating CTkOptionMenu
widgets (e.g., the camera selector dropdown).

Add a monkey-patch that guards against the empty-string return value.
2026-02-22 11:59:51 +02:00
Kenneth Estanislao 07b4d66965 Update version in README to 2.0.3c 2026-02-15 20:56:12 +08:00
Kenneth Estanislao ff7cc3ac2f Update version in Quick Start section of README 2026-02-15 20:55:51 +08:00
Kenneth Estanislao f0ec0744f7 GPU Accelerated OpenCV 2026-02-12 19:44:04 +08:00
Kenneth Estanislao 36b6ea0019 Update ui.py
DETECT_EVERY_N = 2 reuses cached face positions on alternate frames
2026-02-12 18:54:18 +08:00
Kenneth Estanislao 523ee53c34 Update ui.py
Separate capture and processing threads with queue.Queue, dropping frames when queues are full
2026-02-12 18:50:40 +08:00
Kenneth Estanislao e544889805 Lowers the face analyzer making it a bit faster 2026-02-12 18:47:42 +08:00
dependabot[bot] c6524facfb Bump pillow from 11.1.0 to 12.1.1
Bumps [pillow](https://github.com/python-pillow/Pillow) from 11.1.0 to 12.1.1.
- [Release notes](https://github.com/python-pillow/Pillow/releases)
- [Changelog](https://github.com/python-pillow/Pillow/blob/main/CHANGES.rst)
- [Commits](https://github.com/python-pillow/Pillow/compare/11.1.0...12.1.1)

---
updated-dependencies:
- dependency-name: pillow
  dependency-version: 12.1.1
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
2026-02-11 16:36:29 +00:00
Kenneth Estanislao 91baa6c0a5 Update Quick Start section to version 2.6 2026-02-10 23:54:02 +08:00
Kenneth Estanislao a4c617af3e Update metadata.py 2.6 2026-02-10 12:23:28 +08:00
Kenneth Estanislao 9a33f5e184 better mouth mask
better mouth mask showing and tracking the lips part only.
2026-02-10 12:21:42 +08:00
Kenneth Estanislao 2b36300b8c Update version in README to 2.0.2c
- Optimized on video processing with improvements up to 200%
2026-02-06 22:30:39 +08:00