Machine Learning | Audio sounds recognition
Budget: €1,500 – €3,000 EUR
We need to detect the offset of specific sounds from large audios which are NOT words.
Those sounds are always exactly the same, never modified, so they can be labeled. However, they include significant background noise over them.
You will have to build the model and deploy it on AWS under our guidelines.
We need to be able to train the model ourselves creating labels, and processing .mp3 / .wav files as tag sample to train it.
The output of the model must be the % probability of every label per every audio second.
What we need is to own a model on cloud called through API which allows us to use our own files, similar to that this website does with microphone: https://teachablemachine.withgoogle.com/train/audio
We will only pay when one example label is detected correctly. We will provide 100 noise audio samples and 20 label audio samples.
Those sounds are always exactly the same, never modified, so they can be labeled. However, they include significant background noise over them.
You will have to build the model and deploy it on AWS under our guidelines.
We need to be able to train the model ourselves creating labels, and processing .mp3 / .wav files as tag sample to train it.
The output of the model must be the % probability of every label per every audio second.
What we need is to own a model on cloud called through API which allows us to use our own files, similar to that this website does with microphone: https://teachablemachine.withgoogle.com/train/audio
We will only pay when one example label is detected correctly. We will provide 100 noise audio samples and 20 label audio samples.