how does performance change with the length of the audio?

lucko515 / speech-recognition-neural-network

This is the end-to-end Speech Recognition neural network, deployed in Keras. This was my final project for Artificial Intelligence Nanodegree @Udacity.

187 stars 87 forks source link

how does performance change with the length of the audio? #1

Closed StanSilas closed 6 years ago

StanSilas commented 6 years ago

Is this method more suited to shorter bursts of audio or can it be employed for transcription on audio of longer duration say 30-40 minutes? i.e is there a considerable decay in performance when applied to longer audio files?

lucko515 commented 6 years ago

Hmm good point. I haven't had chance to test it out on that long audio. The most samples in the dataset are 1 max 2 sentences.