Go to MPAI-OSD V1.5 AI Modules

Function
Ref. Model
I/O Data
SubAIMs
JSON MData
Profiles
Ref. Software
Conformance
Performance

1 Functions

The Identity Reconciliation (OSD-IDR) AIM reconciles the identifiers that two perception AIMs produce for the same person (the Face ID, from face recognition, and the Speaker ID, from speaker recognition) into a single User ID, and issues the response of the machine to the person.

Receives Face ID An Instance Identifier with the candidate identifiers of the face of the person, ranked with their confidence levels (Port Number 1).
Speaker ID An Instance Identifier with the candidate identifiers of the voice of the person, ranked with their confidence levels (Port Number 2).
Reconciles Face ID and Speaker ID Into one ranked list of candidates: the confidence levels of the two modalities are made comparable and combined per candidate. Where the two modalities agree the confidence is high, where they disagree it is lower. A modality that gives only a coarse label (no candidate matched) does not contribute, and the User ID rests on the other modality.
Produces User ID An Instance Identifier with the reconciled candidate identifiers of the person, the first being the primary one.
Personal Status The Personal Status the machine should display: serious while identifying, welcoming when the person is recognised, reproaching when not.
Response A Basic Text Object with the words the machine says to the person.
Identification A boolean: true when the person has been identified.

User ID, Personal Status and Response are optional outputs; Identification is always produced.

2 Reference Model

Figure 1 depicts the Reference Model of the Identity Reconciliation (OSD-IDR) AIM.

Identity Reconciliation OSD-IDR AIM

Figure 1 – Reference Model of the Identity Reconciliation (OSD-IDR) AIM

3 I/O Data

Table 1 specifies the Input and Output Data of the Identity Reconciliation (OSD-IDR) AIM. Links are to the Data Type specifications.

Table 1 – I/O Data of the Identity Reconciliation (OSD-IDR) AIM

Input Data Type Description
Face ID Instance Identifier (Port Number 1) Candidate identifiers of the person from the face, with their confidence levels.
Speaker ID Instance Identifier (Port Number 2) Candidate identifiers of the person from the voice, with their confidence levels.
Output Data Type Description
User ID Instance Identifier (optional) The reconciled candidate identifiers of the person, ranked.
Personal Status Personal Status (optional) The Personal Status the machine displays.
Response Basic Text Object (optional) The words the machine says.
Identification boolean True when the person has been identified.

4 SubAIMs

No SubAIMs.

5 JSON Metadata

https://schemas.mpai.community/OSD/V1.5/AIMs/IDReconciliation.json

6 Profiles

No Profiles.

7 Reference Software

Not part of this specification.

8 Conformance Testing

Table 2 provides the Conformance Testing Method for the OSD-IDR AIM.

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.

Table 2 – Conformance Testing Method for the OSD-IDR AIM

Receives Face ID (Instance Identifier) Shall validate against the Instance Identifier schema.
Speaker ID (Instance Identifier) Shall validate against the Instance Identifier schema.
Produces User ID (Instance Identifier) Shall validate against the Instance Identifier schema.
Personal Status Shall validate against the Personal Status schema.
Response (Basic Text Object) Shall validate against the Basic Text Object schema.

9 Performance Assessment

Not part of this specification.

Go to MPAI-OSD V1.5 AI Modules