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
Automatic Speech Recognition (MMC-ASR):
Receives | Language Selector |
Auxiliary Text | |
Speech Object | |
Speaker ID | |
Speech Overlap | |
Speaker Time | |
Produces | Recognised Text (also called text transcript). |
Recognised Text can be a Text Segment or just a string.
2 Reference Model
Figure 1 depicts the Reference Model of the Automatic Speech Recognition (MMC-ASR) AIM.
Figure 1 – The Automatic Speech Recognition (MMC-ASR) AIM
3 Input/Output Data
Table 1 specifies the Input and Output Data of the Automatic Speech Recognition (MMC-ASR) AIM.
Table 1 – I/O Data of the Automatic Speech Recognition (MMC-ASR) AIM
Input | Description |
Language Selector | Selects input language |
Auxiliary Text | Text Object with content related to Speech Object. |
Speech Object | Speech Object emitted by Entity |
Speaker ID | Identity of Speaker |
Speech Overlap | Times and IDs of overlapping speech segments |
Speaker Time | Time during which Speech is recognised |
Output | Description |
Recognised Text | Output of the Automatic Speech Recognition AIM, a Text Segment or just a string. |
4 SubAIMs
No SubAIMs.
5 JSON Metadata
https://schemas.mpai.community/MMC/V2.2/AIMs/AutomaticSpeechRecognition.json
6 Profiles
No Profiles
7 Reference Software
7.1 Disclaimers
- This MMM-ASR Reference Software Implementation is released with the BSD-3-Clause licence.
- The purpose of this Reference Software is to demonstrate a working Implementation of MMC-ASR, not to provide a ready-to-use product.
- MPAI disclaims the suitability of the Software for any other purposes and does not guarantee that it is secure.
- Use of this Reference Software may require acceptance of licences from the respective repositories. Users shall verify that they have the right to use any third-party software required by this Reference Software.
7.2 Guide to the ASR code #1
The code takes Speech Objects from MMC-AUS and generates Text Segments (called text transcripts). It uses the whisper-large-v3 model to convert an input Speech Object (speaker’s turn) into a Text Segment (here called text transcript). Disfluencies (e.g., repetitions, repairs, filled pauses) are often omitted. The Whisper reference document is available.
The MMC-ASR Reference Software is found at the MPAI gitlab site. Use of this AI Modules is for developers who are familiar with Python, Docker, RabbitMQ, and downloading models from HuggingFace. The Reference Software contains:
- src: a folder with the Python code implementing the AIM
- Dockerfile: a Docker file containing only the libraries required to build the Docker image and run the container
- requirements.txt: dependencies installed in the Docker image
- README.md: commands for cloning https://huggingface.co/openai/whisper-large-v3
Library: https://github.com/linto-ai/whisper-timestamped
7.3 Guide to the ASR code #2
Use of this AI Modules is for developers who are familiar with Python and downloading models from HuggingFace,
A wrapper for the Whisper NN Module:
- Manages input files and parameters: Speech Object
- Performs Speech Recognition on each Speech Object by executing the Whisper Module.
- Outputs Recognised Text.
The MMC-ASR Reference Software is found at the NNW gitlab site (registration required). It contains:
- The python code implementing the AIM.
- The required libraries are: pytorch and transformers (HuggingFace).
7.4 Acknowledgements
This version of the MMC-ASR Reference Software
- #1 has been developed by the MPAI AI Framework Development Committee (AIF-DC).
- #2 has been developed by the MPAI Neural Network Watermarking Development Committee (NNW-DC).
8 Conformance Testing
Input Data | Data Format | Input Conformance Testing Data |
Speech Object | .wav | All input Speech files to be drawn from Speech files. |
Output Data | Data Format | Output Conformance Testing Criteria |
Recognised Text | Unicode | All Text files produced shall conform with Text files. |
9 Performance Assessment
Performance Assessment of an ASR Implementation (ASRI) can be performed for a language for which there is a dataset of speech segments of various durations with corresponding Transcription Text. An MMC-ASR AIM Performance Assessment Report shall be based on the following steps and specify the input dataset used.
For each Recognised Text produced by the ASRI being Assessed for Performance in response to a speech segment provided as input:
- Compare the Recognised Text with the Transcription Text
- Compute the Word Error Rate (WER) defined as the sum of deletion, insertion, and substitution errors in the Recognised Text compared to the Transcription Text, divided by the total number of words in the Transcription Text.
This code can be used to compute the WER.
Performance Assessment of an ASRI for a language in a Performance Assessment Report is defined as “The WER computed on all speech segments included in the reported dataset”.