1 Function 2 Reference Model 3 Input/Output Data
4 SubAIMs 5 JSON Metadata 6 Profiles
7 Reference Software 8 Conformance Texting 9 Performance Assessment

1 Functions

Audio Separation and Enhancement (CAE-ASE):

Receives Audio Object in the Transform domain.
Audio Spatial Attitudes Spatial Attitudes of the input Audio Objects.
Separates Audio Objects by using their Spatial Attitudes.
Produces Enhanced Audio Object in the Transform domain.
Audio Scene Geometry The Geometry of Audio Objects in the Scene.

2 Reference Model

Figure 1 depicts the Reference Architecture of the Audio Separation and Enhancement AIM.

Figure 1 – Audio Separation and Enhancement AIM

3 Input/Output Data

Table 11 specifies the Input and Output Data of the Audio Separation and Enhancement AIM.

Table 1 – I/O Data of Audio Separation and Enhancement

Input Description
Audio Object The result of the application of the Fast Fourier Transform to the Multichannel Audio.
Audio Spatial Attitudes The Spatial Attitudes of Audio Objects.
Output Description
Enhanced Audio Object Enhanced Multichannel Audio in the transform domain.
Audio Scene Geometry The spatial arrangement of the Audio Objects.

4 SubAIMs

No SubAIMs.

5 JSON Metadata

https://schemas.mpai.community/CAE/V2.3/AIMs/AudioSeparationAndEnhancement.json

6 Profiles

No Profiles.

7 Reference Software

8 Conformance Testing

The following steps shall be followed when testing the Conformance of a CAE-ASE AIM instance.

  1. Use the following datasets:
    1. DS1: n Test files containing Audio Objects (Transform).
    2. DS4: n Test files containing the Spatial Attitudes of Audio Objects.
    3. DS2: n Expected Enhanced Audio Objects (Transform) Files.
    4. DS3: n Expected Audio Scene Geometries.
  2. Feed the AIM under test with the Test Audio Objects (Transform) and Spatial Attitudes.
  3. Analyse the Audio Scene Geometry and Enhanced Audio (Transform).
    1. Control the Audio Scene Geometry with the Expected Audio Scene Geometry:
      1. Count the number of Audio Objects in the Audio Scene Geometry.
      2. Calculate the angle difference (AD) in degrees between the Audio Objects (u) in the Audio Scene Geometry and the Audio Objects (v) in the Expected Audio Scene Geometry.
    2. Compare the number of Audio Blocks in the Expected Audio Objects with the number of Audio Blocks in the Audio Objects (Transform).
    3. Calculate Signal to Interference Ratio (SIR), Signal to Distortion Ratio (SDR), and Signal to Artefacts Ratio (SAR) between the Expected Audio Objects (Transform) and Output Audio Objects (Transform).
    4. Accept the CAE-ASE AIM under test if these four conditions are satisfied:
      1. The number of Audio Objects (Transform) in the Audio Scene Geometry is equal to the number of Audio Objects (Transform) in the Expected Audio Scene Geometry.
      2. The number of Audio Blocks in the Audio Objects (Transform) is equal to the number of Audio Blocks in the Expected Audio Objects (Transform).
      3. Compare each Audio Objects (Transform) in the Audio Scene Geometry with the Audio Objects (Transform) in the Expected Audio Scene Geometry.
        1. Each Audio Objects (Transform)’s AD between the Expected and Output is less than 5 degrees.
      4. Compare each Audio Objects (Transform) with the Audio Objects (Transform) in the Expected Audio Objects (Transform).
        1. If the room reverb time (T60) is greater than 0.5 seconds.
          1. Each object’s SIR between the Expected and Output is greater than or equal to 10 dB.
          2. Each object’s SDR between the Expected and Output is greater than or equal to 3 dB.
          3. Each object’s SAR between the Expected and Output is greater than or equal to 3 dB.
        2. If the room reverb time (T60) is less than 0.5 seconds.
          1. Each object’s SIR between the Expected and Output is greater than or equal to 15 dB.
          2. Each object’s SDR between the Expected and Output is greater than or equal to 6 dB.
  4. The Conformance Tester will provide the following matrix containing a limited number of input records (n) with the corresponding outputs. If an input record fails, the tester would specify the reason why the test case fails.
Input data (DS1, DS4) Expected Output Data (DS2, DS3) Data Format Audio Scene Geometry Source Separation Metrics
Spatial Attitude (ID1)
Audio Object (Transform) ID1		 Enhanced Audio Object (Transform) ID1
Audio Scene Geometry ID1		 T/F T/F T/F
Spatial Attitude (ID2)
Audio Object (Transform) ID2		 Enhanced Audio Object (Transform) ID2
Audio Scene Geometry ID2		 T/F T/F T/F
Spatial Attitude (ID3)
Audio Object (Transform) ID3		 Enhanced Audio Object (Transform ID3
Audio Scene Geometry ID3		 T/F T/F T/F
… … … … …
Spatial Attitude (IDn)
Audio Object (Transform) IDn		 Enhanced Audio Object (Transform IDn
Audio Scene Geometry IDn		 T/F T/F T/F
  1. Final evaluation : T/F Denoting with i, the record number in DS1, DS2, and DS3, the matrices reflect the results obtained with input records i with the corresponding outputs i.
DS1 DS2 DS3 Sound Field Description output value (obtained through the AIM under test)
DS1[i] DS2[i] DS3[i] SpeechDetectionandSeparation[i]

Table 2 provides the Conformance Testing Method for the formats of the CAE-ASE AIM output.

Note: 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 CAE-ASE AIM

Receives Audio Object Shall validate against Audio Object schema.
Audio Data shall conform with Audio Qualifier.
Audio Spatial Attitudes Shall validate against Spatial Attitude schema.
Produces Enhanced Audio Object Shall validate against Audio Object schema.
Audio Data shall conform with Audio Qualifier.
Audio Scene Geometry Shall validate against Audio Basic Scene Geometry schema.

9 Performance Assessment