← Registry / Multimodal Conversation
AIM · composite
Multimodal Access Control
MMC-MAC-V2.5
Submissions for this AIM must match this definition: the same ports, API profile and internal wiring.
Description
Multimodal Access Control: recognises a User from face and speech and issues a spoken/rendered verdict. Face Recognition and Speaker Recognition read the enrolled gallery (the User IDs) and the Scene Descriptors from Shared Storage - these are NOT ports. Time (OSD-STM) accompanies each captured object and is supplied by the User Agent's clock. ID Reconciliation reconciles the two candidate Instance Identifiers and issues the Response (the verdict words) and its Personal Status; Response and Scene Rendering speaks the Response and produces the lip-synced Machine Face Descriptors. The reconciled UserID stays internal; the User consumes the Response and the avatar. Acquisition and delivery are the User Agent, not sub-AIMs.
Inputs and outputs 5 in · 3 out
Inputs
- FaceObject OSD-BVO-V1.5
- BasicSpeechObject OSD-BSO-V1.5
- Prompt optional OSD-BTO-V1.5
- AvatarModel OSD-B3O-V1.5
- MachineID optional string
Outputs
- Response OSD-BTO-V1.5
- Identification boolean
- PortableAvatar PAF-PAV-V1.6
Built from 4 AIMs
| AIM | Code | Instances |
|---|---|---|
| Face Identity Recognition | PAF-FIR-V1.6 | 1 |
| Speaker Identity Recognition | MMC-SIR-V2.5 | 1 |
| Identity Reconciliation | OSD-IDR-V1.5 | 1 |
| Response and Scene Rendering | PAF-RSR-V1.6 | 1 |
Documentation
Implementations 0
None published yet.