Emiliana
|
60fb3aac31
|
Lower overlap threshold.
|
2020-03-25 23:02:23 +01:00 |
|
Emiliana
|
3f85a44f13
|
Slightly higher threshold.
|
2020-03-25 22:57:13 +01:00 |
|
Emiliana
|
9bd3fcc617
|
When multiple faces were close to each other, some of them were removed by accident. This fixes that.
|
2020-03-25 22:52:28 +01:00 |
|
Emiliana
|
5c118347b6
|
New and improved tracking model and confidence value visualization.
|
2020-03-25 17:03:30 +01:00 |
|
Emiliana
|
b0a61b41d5
|
RetinaFace inference is now done at full resolution without padding.
|
2020-03-22 18:01:21 +01:00 |
|
Emiliana
|
381d3b67ec
|
Small bugfix.
|
2020-03-22 14:20:17 +01:00 |
|
Emiliana
|
14b4bd9122
|
As the Haar Cascade face detector is no longer used, it is now remoed.
|
2020-03-22 14:14:26 +01:00 |
|
Emiliana
|
c03e222a49
|
Smaller, faster and better detection model.
|
2020-03-22 13:47:55 +01:00 |
|
Emiliana
|
a10eacc670
|
A new, custom face detection model is now used to lock onto faces for tracking.
|
2020-03-22 00:34:34 +01:00 |
|
Emiliana
|
4f611997f3
|
Fudging thresholds when the head turns left or right is cheating, but it seems to work well.
|
2020-03-08 18:15:43 +01:00 |
|
Emiliana
|
31536929f7
|
These changes further improve point normalization to make it easier to detect expressions under rotations.
|
2020-03-08 18:05:12 +01:00 |
|
Emiliana
|
9c12d29399
|
The OpenSeeVRMDriver example now supports tracking mouth shapes and blinks from camera information.
|
2020-03-08 13:50:53 +01:00 |
|
Emiliana
|
d678f81bda
|
run.bat can be used to run the face tracker visualization in a slightly user friendly way.
|
2020-03-07 20:07:08 +01:00 |
|
Emiliana
|
9aeb4fe95d
|
It is now possible to make certain expressions more important when training SVM models.
|
2020-03-07 13:35:58 +01:00 |
|
Emiliana
|
94d2f4fbe0
|
With this, the face tracker will actually track at the specified FPS, saving CPU especially at lower tracking frame rates.
|
2020-03-03 18:48:52 +01:00 |
|
Emiliana
|
0c84f50672
|
3D points are now normalized by reference sizes.
|
2020-03-01 22:27:50 +01:00 |
|
Emiliana
|
f25f27f1c4
|
The hotkeys weren't actually global, but now they should be.
|
2020-03-01 20:38:47 +01:00 |
|
Emiliana
|
b871090998
|
OpenSeeVRMDriver now can use global hotkeys (Ctrl+Shift+F1 to F6) to override expressions.
|
2020-03-01 19:29:17 +01:00 |
|
Emiliana
|
a8648eeb80
|
Some small fixes.
|
2020-03-01 19:26:45 +01:00 |
|
Emiliana
|
5059ef9cbf
|
The overriding of point selection from models is now disabled.
|
2020-03-01 14:40:30 +01:00 |
|
Emiliana
|
04d7f5b8b3
|
This adds manually designed features to the OpenSeeData.
|
2020-03-01 14:23:56 +01:00 |
|
Emiliana
|
7706372a7a
|
This fixes exceptions with network tracking.
|
2020-02-28 20:34:39 +01:00 |
|
Emiliana
|
04e19c9798
|
This adds optimizations and safety checks to the 3D point adjustments and adds an option to the IK target, which allows its position to slowly drift back to neutral.
|
2020-02-28 20:19:03 +01:00 |
|
Emiliana
|
033a48b132
|
This changes the points used for PnP solving and replaces the RetinaFace model with a higher resolution version.
|
2020-02-28 02:10:23 +01:00 |
|
Emiliana
|
4a88c57be2
|
This fixes some bugs.
|
2020-02-27 00:38:46 +01:00 |
|
Emiliana
|
cd9cfd9ee8
|
RetinaFace was added as a face detection scheme.
|
2020-02-26 00:54:32 +01:00 |
|
Emiliana
|
90c7c9284b
|
The datasets used for training the new gaze tracker are now mentioned in the README file.
|
2020-02-24 23:31:35 +01:00 |
|
Emiliana
|
44060dfff2
|
The eye tracking model was switched out. This one seems to be more stable.
|
2020-02-24 21:51:54 +01:00 |
|
Emiliana
|
e1a3b69d22
|
Some bugs in the gaze applier were fixed.
|
2020-02-24 01:22:23 +01:00 |
|
Emiliana
|
0332f9cb93
|
This adds the gaze tracker.
|
2020-02-24 00:44:17 +01:00 |
|
Emiliana
|
2d35cc2aaf
|
OpenSeeVRMDriver can now also apply blinks from face tracking instead of blinking automatically.
|
2020-02-21 18:52:11 +01:00 |
|
Emiliana
|
6c7392b037
|
Learning expression actually should not depend on the location of the head. This fixes that issue.
|
2020-02-21 01:05:41 +01:00 |
|
Emiliana
|
e60b9b37bd
|
The internal face 3D model will now adapt to the tracked landmarks.
|
2020-02-20 21:33:00 +01:00 |
|
Emiliana
|
7de50b2e25
|
When using OpenSeeLauncher with pinvoke, it will now log process output to log files.
|
2020-02-17 20:17:34 +01:00 |
|
Emiliana
|
e157e9d5c9
|
Process.Start is not supported in IL2CPP builds, so now OpenSeeLauncher has an option to do the process launching directly through DLL calls.
|
2020-02-17 00:25:25 +01:00 |
|
Emiliana
|
633293eb52
|
The face tracker can now offset face ids by a given value. This makes it possible to easily mix tracking input from multiple network sources.
|
2020-02-16 15:58:47 +01:00 |
|
Emiliana
|
c2eb544096
|
The UDP socket is now actually closed when the OpenSee component is destroyed.
|
2020-02-16 01:46:35 +01:00 |
|
Emiliana
|
05a9be7b71
|
Mono and stereo audio should now both be handled correctly for lip sync. More than 2 channels are not supported.
|
2020-02-15 14:58:22 +01:00 |
|
Emiliana
|
200b182ab9
|
Switching back from canned audio to microphone for lip sync in OpenSeeVRMDriver didn't work. This fixes that.
|
2020-02-15 13:44:54 +01:00 |
|
Emiliana
|
626ce2a1ae
|
Two flags were added to OpenSeeVRMDriver to allow reinitialization of different parts.
|
2020-02-15 13:35:10 +01:00 |
|
Emiliana
|
2c08712aef
|
The OpenSeeVRMDriver example uses OVRLipSync to do lip syncing from audio and apply it to a VRM avatar, while also handling expressions detected using OpenSeeExpression and applying eye blinks.
|
2020-02-15 13:21:46 +01:00 |
|
Emiliana
|
bb845098ca
|
It is now possible to train expression models with probability support. This however is a lot slower! Also, the C parameter for SVM training was set incorrectly. Now training works much better.
|
2020-02-13 22:24:41 +01:00 |
|
Emiliana
|
9abc3029e7
|
OpenSeeExpression now supports detecting expressions on point subsets.
|
2020-02-13 21:15:19 +01:00 |
|
Emiliana
|
de16fc9ff9
|
The IK target component can now mirror the pose, which can be convenient.
|
2020-02-13 20:18:05 +01:00 |
|
Emiliana
|
1f3bf30166
|
There's now an example on how to apply detected expressions to a VRM model.
|
2020-02-11 09:55:31 +01:00 |
|
Emiliana
|
443a7ee88e
|
The number of samples for expression learning is now 1000 per expression, because gathering the maximum number of samples is no longer enforced.
|
2020-02-10 21:46:01 +01:00 |
|
Emiliana
|
e83e965290
|
Reading from standard input should now work correctly.
|
2020-02-10 20:26:46 +01:00 |
|
Emiliana
|
b5218a1eb0
|
Training with less data is now possible too.
|
2020-02-09 18:09:09 +01:00 |
|
Emiliana
|
b5706183c2
|
Incomplete expressions are now skipped.
|
2020-02-09 15:00:26 +01:00 |
|
Emiliana
|
750f9920e5
|
OpenSeeIKTarget is now has tooltips and can interpolate between frames, making motion much smoother. Another coordinate issue was fixed too. OpenSeeExpression can now record additional calibration data for fully recorded expressions.
|
2020-02-09 14:34:50 +01:00 |
|