1.
Q: Why can not my webcam be used after updating?
A: It is because of api-ms-win-downlevel-shlwapi-l1-1-0.dll missing.
For win7: needing update this package
https://www.microsoft.com/en-nz/download/details.aspx?id=36805
For win10: update to last version.

2.
Q: Why the avatar's gaze going to upside or downside when irises tracking?
A: For now with irises sync, the webcam position must be right on front face. This problem will be fixed on furture.

3.
Q: How to start hands sync?
A: Once you select the target color of hands tracking, the sync should be starting.

Update announcement:




beta 1.1.1
Add left and right hands can be tracked with different colors.

beta 1.1.0
1. Added irises tracking.
2. Added simplified mode for hand tracking with color as the tracking target, bright gloves and plain background will give better results. Simplified mode is without fingers tracking.
3. Added voice input, sync to the model's lips.
4. Added enhanced face tracking stability algorithm.
5. Added enhanced face tracking robust algorithm with glasses.
All of the above require extra calculations, please choose according to your needs.
6.VRM SDK updated to 0.48.
7. Unity updated to 2018.2.20, does not affect compatibility, no need to re-import the models.
8. Other miscellaneous items fixed.

Finally be returning to the update, wish guys have a good time this year too.

Update announcement:
1.0.1c
Fixed screen error of non-display webcam
Improved face tracking mapping algorithm
1.0.1b
Hotfix: vrm sync bugs
1.0.1
In the past version, wakaru was not well synchronized with the model. It looks a bit stuck, and some people have a very serious laggy.
So rewrote the synchronization calculation this time.
Greatly improved the stability and fluency of face tracking synchronization.
Also reduce CPU load

Update announcement:

In the process of writing the manual,
found that VRM SDK has some problem of the compatibility with converting from fbx to vrm of unity 2017 .
Although it does not affect the VRM those had been successfully converted
But to avoid potential compatibility issues and to unify the version,
I think its time to updated engine version to Unity 2018.2.1f1.

The assetbundle build by Unity 2017 can no longer be used.
Must build assetbundle again by using Unity 2018.
The building steps is the same.
If this makes you feeling annoying, please forgive me.

Update announcement:

1. Added support for VRM model importing.
Put your .vrm in the following path
(your path)\wakaru\Models\VRM\ (your .vrm)
The manual of how to modify simpily to VRM will be added on the official website in a few days.
Includes adding emoticons and adding edge line of model.

2. The import of the model steps in unity no longer checks the facial morphs.
The following morphs lists will be automatically corresponded to facial tracking.
If something missing, this will not cause any errors , just not active the corresponding facial tracking .
MMD is the corresponding name with
eyes:["まばたき", "笑い", "なごみ"]
brows:[ "上", "怒り", "にこり", "真面目", "困る", "寄せ", "下" ]
lips:["あ", "い", "う", "え", "お"]
VRM is the corresponding presets with
eyes:["Blink"]
lips:["A",”I”,”U”,”E”,”O”]

3. VRM is not doing classification of facials, so adding the function of that
the facial UI can be dragged to change the category with mouse.

4. Added automatic blink function.

5. Added console for Debug. Press "~" to show up.

Update announcement:
Add the sensitivity adjustment of facial tracking:
1. Sync Thresholds - Filters tiny shake and drastic changes of facial contour tracking to make avatar motions more stable.
2. Smoothing Settings - If the motions of your avatar are lagging. This can be set the computing and interpolation
3. Head Pose Control - If not in tracking mode, it can be control with mouse.

Reservation Features:
VRM format support
save the settings

Update announcement:
Added the function to use the video as a facial tracking input. The recommended usage is:
You can use webcam to record video and then use the video input as a tracking input to add expressions and motions.
Or any possible usage.

Because face detection does matrix calculations for all input pixels,
The higher the resolution of the video, the higher loading will be. The input resolution not allowed over 1280*720.
It is best to do pre-processing of resolution, grayscale and contrast, and brightness adjustment for recorded videos.

Recommended resolution: 853*480 ~ 640*480, 30fps
Support encode: h.264, vp8

Next update will add the sensitivity adjustment of facial tracking.

Please check the manual on steam store web right side.
You can significantly improve the facial tracking accuracy by adjusting your webcam.
* 1. Set to grayscale.
* 2. Adjust brightness and contrast.
The accuracy of facial tracking so dependents on the user's environment, including
* 1. Don't wear glasses.
* 2. Room lighting must be bright.