SvenWientjes/SpeechRecognition

Classifying sound signals as Links, Midden or Rechts using features computed using a Mel-Frequency filterbank, summing the power of the frequency-domain in the relevant filters. Dynamic Time Warping is used to find proper alignment between the unknown word and several labelled exemplars per word we are looking for. Then, k nearest neighbours tells us which is the most likely class for our unknown word.

21
/ 100
Experimental

This helps researchers or students in linguistics or signal processing classify short, distinct spoken words. You provide sound files of specific words, and it outputs a classification for an unknown word, along with visual plots of how it matched the sounds. This is ideal for those studying word recognition in small vocabularies.

No commits in the last 6 months.

Use this if you need to classify a very limited set of spoken words, like 'Links', 'Midden', or 'Rechts', from audio samples.

Not ideal if you need to recognize a large vocabulary of words, process continuous speech, or require real-time recognition.

speech-recognition linguistics-research signal-processing audio-classification phonetics-analysis
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 4 / 25
Maturity 8 / 25
Community 9 / 25

How are scores calculated?

Stars

7

Forks

1

Language

Matlab

License

Last pushed

Jul 24, 2018

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/SvenWientjes/SpeechRecognition"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.