torzdf
|
3e6b78e9bb
|
Add gcc/g++ checks for linux. Add more informative output
|
2018-11-16 18:24:16 +00:00 |
|
torzdf
|
d1705b08cd
|
Improve setup.py for Windows
|
2018-11-16 17:31:27 +00:00 |
|
torzdf
|
ca63242996
|
Extraction - Speed improvements (#522) (#523)
* Extraction - Speed improvements (#522)
* Initial Plugin restructure
* Detectors to plugins. Detector speed improvements
* Re-implement dlib aligner, remove models, FAN to TF. Parallel processing
* Update manual, update convert, implement parallel/serial switching
* linting + fix cuda check (setup.py). requirements update keras 2.2.4
* Add extract size option. Fix dlib hog init
* GUI: Increase tooltip width
* Update alignment tool to support new DetectedFace
* Add skip existing faces option
* Fix sort tool to new plugin structure
* remove old align plugin
* fix convert -skip faces bug
* Fix convert skipping no faces frames
* Convert - draw onto transparent layer
* Fix blur threshold bug
* fix skip_faces convert bug
* Fix training
|
2018-10-27 10:12:08 +01:00 |
|
1132719438
|
1ce5214e59
|
Fixed face extract with dlib on OS X (#510)
* fix: gpu stat for OS X
- Use pynvx to get GPU info on OS X
- fix face extract with dlib
Issue #488
|
2018-10-01 11:38:59 +01:00 |
|
Ruggero Turra
|
6f004e4b24
|
fix typo (#406)
|
2018-05-19 00:27:38 +01:00 |
|
DKing
|
7ab5a56bff
|
Fix bug in docker image tf/tf:latest-gpu-py3 (#400)
|
2018-05-14 18:42:34 +01:00 |
|
DKing
|
431ff543d5
|
Unified Dependency Solution (#370)
* Unified install helper
smart load dep from requirements-docker.txt
working GUI in docker
compile dlib with avx&cuda
* improve tf version selection
|
2018-05-10 19:39:11 +01:00 |
|