1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
|
## TaSTT: A deliciously free STT
TaSTT (pronounced "tasty") is a free speech-to-text tool for VRChat. It uses
local machine translation to turn your voice into text, then sends it into
VRChat via OSC. A few parameters, a machine-generated FX layer, and a
custom shader display the text in game.
Features:
* 8x22 display grid, 80 characters per slot.
* Text-to-text interface.
* Speech-to-text interface (planned)
* Free as in beer.
* Free as in freedom.
* Hackable.
* 100% from-scratch implementation.
* Permissive MIT license.
Contents:
1. [Motivation](#motivation)
2. [Design overview](#design-overview)
3. [Contributing](#contributing)
4. [Backlog](#backlog)
Made with love by yum\_food.
### Motivation
Many VRChat players choose not to use their mics, but as a practical matter,
occasionally have to communicate. I want this to be as simple, efficient, and
reliable as possible.
There are existing tools which help here, but they are all imperfect for one
reason or another:
1. RabidCrab's STT costs money and relies on cloud-based translation. I have
struggled with latency, quality, and reliability issues. It's also
closed-source.
2. The in-game text box is only visible to your friends list, making it
useless for those who like to make new friends.
Thus I believe that a free alternative is both needed and justified.
I hope that this codebase aids and motivates the creation of better, more
expressive communication tools for mutes.
### Design overview
There are currently 4 important pieces:
1. `TaSTT.shader`. A simple unlit shader. Has one parameter per cell in the
display.
2. `generate\_animations.sh`. Generates one animation per (row, column, letter).
These animations allow us to write the shader's parameters from an FX layer.
3. `generate\_fx.py`. Generates a colossal FX layer which maps (row, column,
letter, active) to exactly one of TaSTT.shader's parameters.
4. `osc\_ctrl.py`. Sends OSC messages to VRChat, which it dutifully passes along
to the generated FX layer.
#### Parameters & board indexing
There are 2 obvious ways to tell the board how to display a message:
1. Independently parameterize every character slot. If we want to display
a 140-character tweet, this means using (140 characters) * (8 bits
per character) == 1120 bits of parameter memory. VRChat only gives us 256!
2. Parameterize one character slot. We could have an 8-bit letter, an 8-bit row
select, and an 8-bit column select. To avoid overwriting cells while we seek,
we could include a 1-bit enable. This approach works, and uses very few
parameter bits, but it requires us to update the same parameter very quickly.
Experimental results with this were not promising; remote viewers would see
the wrong letters pretty often.
Thus I settled on a hybrid approach: we divide the board into `cells`,
inside of which we can independently address each character slot. There
are currently 16 cells.
Since the board has (22 columns) * (8 rows) == 176 character slots, each cell
contains (176 characters) / (16 cells) = 11 characters.
To update a cell, we do this for every single character:
1. Select the cell. Since there are 16 cells, this requires 4 bits.
2. Select the letter. Since we support 256 letters per cell, this requires 8 bits.
To avoid overwriting cells while we seek around, we also have a single boolean
which enables/disables updating any cells.
Thus the total amount of parameter memory used is dictated by this equation:
```
ROWS * COLS * (log2(CELLS) + 8) / CELLS + 1
```
This is currently 133 bits.
#### FX controller design
The FX controller (AKA animator) is pretty simple. There is one layer for each
character in a cell. Thus the layer has to work out which cell it's in, then
work out which letter we want to write in that cell, then run an animation for
that letter.
Here's a layer where I manually moved things around to show the structure of
the decision tree:
### Contributing
Contributions welcome. Send a pull request to this repository.
To use the STT:
1. Enable Windows Subsystem for Linux. This is a lightweight Linux virtual
machine that runs on your Windows host. You can access the Windows
filesystem at /mnt/c/....
2. $ cd /mnt/c/path/to/your/unity/project
2. $ cd Assets
3. $ git clone https://github.com/yum\_food/TaSTT
4. $ cd TaSTT
5. $ ./generate.sh
6. Put TaSTT\_fx.controller and TaSTT\_params.asset on your avatar.
7. Upload (or build & test).
8. Open powershell.
9. Navigate to TaSTT.
10. $ python3 ./osc\_ctrl.py
11. Start typing. Your messages should show display in-game.
### Backlog
1. Better Unity integrations
1. Port all scripts to Unity-native C# scripts.
2. Support appending to existing FX layers.
3. Use VRCSDK to generate FX layer instead of generating the serialized files.
4. Optimize FX layer. Unity takes quite a while to load in the current one.
Some redesign is likely needed.
2. In-game usability features.
1. Resizing (talk to friends far away).
2. Basic toggles (hide it when not needed).
3. World mounting (leave it in a fixed position in world space).
4. Avatar mounting (attach it to your hand).
5. Controller triggers (avoid having to use the radial menu every time you
want to speak).
3. General usability features.
1. Error detection & correction.
2. Text-to-text interface. Type in terminal, show in game.
4. Optimization
1. Utilize the avatar 3.0 SDK's ability to drive parameters to reduce the
total # of parameters (and therefore OSC messages & sync events). Note
that the parameter memory usage may not decrease.
5. Bugfixes
|
