Data

Tools

Indexes

Applications

Experiments

Open Source Science

https://github.com/bagustris/sner

SNER = Speech Naturalness and Emotion Recognition

https://github.com/bagustris/sner

Science Score: 39.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ✓
    DOI references
    Found 1 DOI reference(s) in README
  • ○
    Academic publication links
  • ○
    Committers with academic emails
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (10.9%) to scientific vocabulary
Last synced: 10 months ago · JSON representation

Repository

SNER = Speech Naturalness and Emotion Recognition

Basic Info
  • Host: GitHub
  • Owner: bagustris
  • Language: Python
  • Default Branch: master
  • Size: 60.8 MB
Statistics
  • Stars: 2
  • Watchers: 1
  • Forks: 0
  • Open Issues: 0
  • Releases: 1
Created almost 5 years ago · Last pushed 12 months ago
Metadata Files
Readme

README.md

sner

SNER = Speech Naturalness and Emotion Recognition. SNER is a very simple inference of dimensional speech emotion (valence, arousal, dominance) and naturalness using a machine learning pre-trained model. The program focuses on light size and fast inference time. The model is saved in JOBLIB format (60MB) with an inference time of less than one second (0.8s inference time for 13s audio file).

Input-output format

Input: speech file (e.g., wav) readable by audiofile package
Output: Score of valence, arousal, dominance, and naturalness in the range [-1, 1].

Installation

install dependencies and use the software inside the parent directory.

python3 -m pip install -r requirements.txt

Usage

python3 predict_vadn.py input.wav

or (need to: chmod +x predict_vadn.py)

./predict_vadn.py input.wav

Arguments

bash Positional (required): 'input file' (wav, ogg, mp3) Optional: -m 'path', path of pre-trained model in a JOBLIB format -s 'split', 'chunks' (every duration seconds), or 'full' (without split) -d 'duration', duration (in seconds) if split chunks, default duration=10 -h, show the help

Example

```

to predict each 2-second duration of audio file

bagus@m049:snerosfull$ ./predictvadn.py bagus-test16000.wav -d 2 Valence, arousal, dominance, naturalness #0: [[-0.09434319 0.44684726 -0.08786711 0.09021541]] Valence, arousal, dominance, naturalness #1: [[-0.14146665 0.6224453 -0.19895521 0.13970129]] Valence, arousal, dominance, naturalness #2: [[-0.06549488 0.42078984 -0.07323465 0.14856477]] Valence, arousal, dominance, naturalness #3: [[-0.09165932 0.6154841 -0.1972353 0.2185024 ]] Valence, arousal, dominance, naturalness #4: [[-0.07417645 0.41887206 -0.05865883 0.22063532]] Valence, arousal, dominance, naturalness #5: [[-0.04649594 0.53053147 -0.11787422 0.19543049]] Valence, arousal, dominance, naturalness #6: [[-0.12129027 0.48048788 -0.11438508 0.14086896]] Valence, arousal, dominance, naturalness #7: [[-0.07501961 0.50562567 -0.12277649 0.08429483]] Valence, arousal, dominance, naturalness #8: [[-0.1845332 0.47495478 -0.1136996 0.09605219]]

to predict the whole duration in a single value for each variable

bagus@m049:snerosfull$ ./predictvadn.py bagus-test16000.wav -s full Valence, arousal, dominance, naturalness: [[-0.1591546 0.37833244 -0.06329431 0.39182937]] ```

Demo (version 1.0)

YouTube: https://youtu.be/doZbrVsPpSU

asciicast

Citation

Parts of the software are used in and based on the following paper. Please cite this paper if you use this software. B. T. Atmaja, A. Sasou, and M. Akagi, “Speech Emotion and Naturalness Recognition with Multitask and Single-task Learnings,” IEEE Access, pp. 1–1, 2022, doi: 10.1109/ACCESS.2022.3189481.

License and Contact

The license of the software is PolyForm Noncommercial License 1.0.0; see the attached file. The software is provided as it is without any warranty. It is free for academic and research purposes but prohibited for commercial use. For commercial and other questions, contact me at bagustris@outlook.com.

Owner

  • Name: Bagus Tris Atmaja
  • Login: bagustris
  • Kind: user
  • Location: Tsukuba
  • Company: AIST

Researcher @aistairc @VibrasticLab

GitHub Events

Total
  • Push event: 1
Last Year
  • Push event: 1

Committers

Last synced: about 1 year ago

All Time
  • Total Commits: 31
  • Total Committers: 1
  • Avg Commits per committer: 31.0
  • Development Distribution Score (DDS): 0.0
Past Year
  • Commits: 0
  • Committers: 0
  • Avg Commits per committer: 0.0
  • Development Distribution Score (DDS): 0.0
Top Committers
Name Email Commits
Bagus Tris Atmaja b****s@y****m 31

Issues and Pull Requests

Last synced: about 1 year ago

All Time
  • Total issues: 0
  • Total pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Total issue authors: 0
  • Total pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 0
  • Pull requests: 0
  • Average time to close issues: N/A
  • Average time to close pull requests: N/A
  • Issue authors: 0
  • Pull request authors: 0
  • Average comments per issue: 0
  • Average comments per pull request: 0
  • Merged pull requests: 0
  • Bot issues: 0
  • Bot pull requests: 0
View more stats
Top Authors
Issue Authors
Pull Request Authors
Top Labels
Issue Labels
Pull Request Labels

Dependencies

requirements.txt pypi
  • audiofile >=0.4.1,<=1.0.3
  • joblib ==1.0.1
  • opensmile ==2.4.1
  • scikit-learn ==1.0.2