Function
Ref. Model
I/O Data
SubAIMs
JSON MData
Profiles
Ref. Software
Conformance
Performance
1 Functions
The A‑User Formation (PGM‑AUF) AIM forms the Portable Avatar that utters the A‑User’s response. It combines the deliberated communicative content (the Final Response produced by Basic Knowledge: the text, the Action Objects and the Process Actions) with the A-User State produced by Personality Alignment. The Portable Avatar expresses the A‑User’s Personal Status across speech, face, and body. A-User Formation is driven by the AUC-AUF Request of A-User Control. Its function is that of Response and Scene Rendering (PAF-RSR) of MPAI-PAF, to which it adds the Request and the Response of A-User Control and the division of the Process Actions.
A-User Formation sends nothing to the M-Instance and nothing to the User Agent. It gives the Portable Avatar, and the Process Actions that do not concern the A-User as an entity, to A-User Control, which decides what leaves the A-User.
The A‑User State conveys the full Entity State of the A‑User; the current A‑User Formation SubAIMs consume only its Personal Status component, since the speech, face, and body rendering SubAIMs operate on Personal Status. The interface carries the complete Entity State so that future SubAIMs may exploit more of it without an interface change.
A‑User Formation operates through a single operative interface:
Operative Interface — receives an AUC-AUF Request from A‑User Control and returns an AUC-AUF Response. The AUC-AUF Request controls the formation by element (Speech, Face, Body and Avatar), sets the spatial output, and carries the references to the Avatar Model and the Speech Model to be read from Private Storage. The AUC-AUF Response reports the outcome of each element and of the Process Actions concerned.
PGM‑AUF:
- Receives the AUC-AUF Request from A-User Control, the Avatar Model and the Speech Model read from Private Storage, the Final Response (with its Process Actions) from Basic Knowledge, and the A-User State from Personality Alignment.
- Forms the Portable Avatar that utters the Final Response, expressing the Personal Status component of the A-User State across speech, face, and body, and moving its face and body as the Action Objects and the Process Actions require.
- Produces the Portable Avatar, the Process Actions that do not concern the A‑User as an entity and the AUC-AUF Response, all given to A‑User Control.
Division of the Process Actions. A-User Formation examines each Process Action of the Final Response. If its subject is the Persona of the A-User, that is, the avatar (its position, its movement, its gestures), the Process Action concerns the A-User as an entity and A-User Formation turns it into Motion Objects. If its subject is another Item or Process (for example, a door to be opened), A-User Formation passes the Process Action to A-User Control unchanged. Every Process Action goes one way or the other, none both. The AUC-AUF Response states the two groups.
Motion. A-User Formation derives the Face Motion Objects and the Body Motion Objects that animate the Portable Avatar. It derives them from each Action Object and, by splitting each Process Action that concerns the A-User as an entity into a sequence of Motion Objects (mini-PAs), from the Process Actions. The Motion Objects are expressed in a vocabulary that their Qualifier names. A-User Formation reports the progress of the Process Actions concerned in the AUC-AUF Response, with the PA Status, and stops a motion when the AUC-AUF Request so requests. A-User Formation never sends a Process Action to the M-Instance: the Process Actions are sent by A-User Control, if it lets them through.
2 Reference Model
Figure 1 depicts the Reference Model of the A‑User Formation (PGM‑AUF) AIM.
Figure 1 – Reference Model of the A‑User Formation (PGM‑AUF) AIM
3 I/O Data
Table 1 specifies the Input and Output Data of the A‑User Formation (PGM‑AUF) AIM.
| Input | Description |
|---|---|
| AUC-AUF Request | Request from A-User Control that controls the formation by element (Speech, Face, Body and Avatar), sets the spatial output, and includes the references to the Avatar Model and Speech Model to be read from Private Storage. |
| Avatar Model | The A‑User’s avatar model, read from Private Storage. |
| Speech Model | The A‑User’s speech (voice) model, read from Private Storage. |
| Final Response | The deliberated content produced by Basic Knowledge: the text to be uttered, the Action Objects that carry the actions of the face and body that accompany it and the Process Actions, of which the ones that concern the A-User as an entity are turned into Motion Objects. |
| A‑User State | The Entity State the A‑User should adopt, produced by Personality Alignment; its Personal Status component drives expressive rendering. |
| Output | Description |
| Portable Avatar | The synthesised Portable Avatar expressing the A-User’s Personal Status across speech, face, and body, given to A-User Control, which decides whether it leaves the A-User. |
| Process Actions | The Process Actions, in the syntax of MMM-TEC V2.2, that do not concern the A-User as an entity, passed unchanged to A-User Control. |
| AUC-AUF Response | Response describing the execution and outcome of avatar formation and the progress of the Process Actions concerned (PA Status), reported to A-User Control. |
4 SubAIMs (informative)
This section is informative. The decomposition into SubAIMs described below illustrates one conformant architecture for producing the normative outputs of PGM‑AUF. Implementations may adopt alternative internal structures provided they satisfy the conformance requirements of Section 8.
4.1 Reference Model
Figure 2 depicts the Reference Model of the A‑User Formation (PGM‑AUF) Composite AIM.
Figure 2 – Reference Model of the A‑User Formation (PGM‑AUF) Composite AIM
4.2 Operation
The AUF operation includes the following SubAIMs:
- Personal Status Demultiplexing — extracts the Personal Status component from the A‑User State and splits it into Speech, Face, and Body Personal Status.
- Text-To-Speech — synthesises Machine Speech from the Final Response and the Speech Personal Status, using the Speech Model (an MPAI‑MMC AIM).
- Motion Object Generation — divides the Process Actions of the Final Response: those that concern the A‑User as an entity (their subject is the Persona of the A‑User) become Face Motion Objects and Body Motion Objects, together with those derived from the Action Objects; the others are passed to A‑User Control.
- Entity Face Description — produces the Face Descriptors from the Machine Speech, mainly, and from the Final Response, the Face Personal Status and the Face Motion Objects.
- Entity Body Description — produces the Body Descriptors from the Final Response, the Body Personal Status and the Body Motion Objects.
- Portable Avatar Multiplexing — multiplexes the Avatar Model, the Machine Speech, the Face Descriptors, and the Body Descriptors into the Portable Avatar (an MPAI‑PAF AIM).
- AUC-AUF Response Multiplexing — combines the AUC-AUF Response of Personal Status Demultiplexing, Entity Face Description, Entity Body Description and Motion Object Generation into the AUC-AUF Response reported to A‑User Control.
The AUC-AUF Request is received by every SubAIM that produces the Portable Avatar or its components, namely Personal Status Demultiplexing, Text-To-Speech, Motion Object Generation, Entity Face Description, Entity Body Description and Portable Avatar Multiplexing, so that A‑User Control can influence each of them.
4.3 Functions of SubAIMs
Table 2 specifies the functions of the A‑User Formation (PGM‑AUF) SubAIMs.
| SubAIM | Function |
|---|---|
| Personal Status Demultiplexing | Extracts the Personal Status from the A‑User State and splits it into Speech, Face, and Body Personal Status. |
| Text-To-Speech | Synthesises Machine Speech from the Final Response and the Speech Personal Status, using the Speech Model. |
| Motion Object Generation | Divides the Process Actions of the Final Response; generates Face Motion Objects and Body Motion Objects from the Process Actions that concern the A-User as an entity and from the Action Objects; passes the other Process Actions to A-User Control. |
| Entity Face Description | Produces Face Descriptors mainly from the Machine Speech, and from the Final Response, Face Personal Status and Face Motion Objects. |
| Entity Body Description | Produces Body Descriptors from the Final Response, Body Personal Status and Body Motion Objects. |
| Portable Avatar Multiplexing | Multiplexes the Avatar Model, Machine Speech, Face Descriptors, and Body Descriptors into the Portable Avatar. |
| AUC-AUF Response Multiplexing | Combines the AUC-AUF Response of the SubAIMs into the AUC-AUF Response of the AIM. |
4.4 I/O Data of SubAIMs
Table 3 gives the Input and Output Data of the A‑User Formation (PGM‑AUF) SubAIMs.
| SubAIM | Input | Output |
|---|---|---|
| Personal Status Demultiplexing | A‑User State AUC-AUF Request |
Speech Personal Status Face Personal Status Body Personal Status AUC-AUF Response |
| Text-To-Speech | AUC-AUF Request Speech Model Final Response Speech Personal Status |
Machine Speech |
| Motion Object Generation | AUC-AUF Request Final Response |
Face Motion Object Body Motion Object Process Actions AUC-AUF Response |
| Entity Face Description | AUC-AUF Request Final Response Machine Speech Face Motion Object Face Personal Status |
Face Descriptors AUC-AUF Response |
| Entity Body Description | AUC-AUF Request Final Response Body Motion Object Body Personal Status |
Body Descriptors AUC-AUF Response |
| Portable Avatar Multiplexing | Avatar Model Machine Speech Face Descriptors Body Descriptors AUC-AUF Request |
Portable Avatar |
| AUC-AUF Response Multiplexing | AUC-AUF Response (from Personal Status Demultiplexing) AUC-AUF Response (from Entity Face Description) AUC-AUF Response (from Entity Body Description) AUC-AUF Response (from Motion Object Generation) |
AUC-AUF Response |
4.5 AIMs and JSON Metadata
Table 4 provides the links to the AIM specifications and JSON schemas. AIM1 indicates the Composite AIM and AIM2 its SubAIMs.
| AIM1 | AIM2 | Name | JSON |
|---|---|---|---|
| PGM‑AUF | A‑User Formation | X | |
| PGM‑PSD | Personal Status Demultiplexing | X | |
| MMC‑TTS | Text-To-Speech | X | |
| PGM‑MOG | Motion Object Generation | X | |
| PGM‑EFD | Entity Face Description | X | |
| PGM‑EBD | Entity Body Description | X | |
| PAF‑PMX | Portable Avatar Multiplexing | X | |
| PGM‑FSM | AUC-AUF Response Multiplexing | X |
5 JSON Metadata
https://schemas.mpai.community/PGM1/V1.0/AIMs/AUserFormation.json
6 Profiles
No Profiles.
7 Reference Software
Not part of this specification.
8 Conformance Testing
Table 5 provides the Conformance Testing Method for the A‑User Formation (PGM‑AUF) AIM. Conformance Testing of the individual SubAIMs is given by the individual AIM specifications.
If a schema contains references to other schemas, conformance of data for the primary schema implies that any data referencing a secondary schema shall also validate against the relevant schema, if present, and conform with the Qualifier, if present.
| Receives | AUC-AUF Request | Shall validate against AUC-AUF Request schema. |
| Avatar Model | Shall validate against Avatar Model schema. | |
| Speech Model | Shall validate against Speech Model schema. | |
| Final Response | Shall validate against Final Response schema. | |
| A‑User State | Shall validate against A‑User State schema. | |
| Produces | Portable Avatar | Shall validate against Portable Avatar schema. |
| Process Actions | Shall validate against the Process Actions schema and shall not include a Process Action turned into Motion Objects. | |
| AUC-AUF Response | Shall validate against AUC-AUF Response schema. |
9 Performance Assessment
Not part of this specification.
