Hello everyone,\nI have a little problem. Indeed, I don't know how I can test on the test dataset, because I didn't understand how to use the two audio files. My model uses spectrograms, so I take each audio extract and transform it into a spectrogram. \nHowever, I don't understand how to do this on the two audio files. Should I divide them into small clips and then turn them into a spectrogram and then predict which bird sings in each clip, or should I do something else?
\n\nPS: Should I download a test dataset? (I saw on a notebook that it had a dataset with lots of little clips, is that what I'm looking for)?
", "messages": [ { "id": "901396", "postDate": "06/25/2020 12:48:25", "content": "Hello everyone,\nI have a little problem. Indeed, I don't know how I can test on the test dataset, because I didn't understand how to use the two audio files. My model uses spectrograms, so I take each audio extract and transform it into a spectrogram. \nHowever, I don't understand how to do this on the two audio files. Should I divide them into small clips and then turn them into a spectrogram and then predict which bird sings in each clip, or should I do something else?
\n\nPS: Should I download a test dataset? (I saw on a notebook that it had a dataset with lots of little clips, is that what I'm looking for)?
", "rawMarkdown": "Hello everyone,\nI have a little problem. Indeed, I don't know how I can test on the test dataset, because I didn't understand how to use the two audio files. My model uses spectrograms, so I take each audio extract and transform it into a spectrogram. \nHowever, I don't understand how to do this on the two audio files. Should I divide them into small clips and then turn them into a spectrogram and then predict which bird sings in each clip, or should I do something else?\n\nPS: Should I download a test dataset? (I saw on a notebook that it had a dataset with lots of little clips, is that what I'm looking for)?", "votes": null }, { "id": "975238", "postDate": "08/18/2020 08:00:57", "content": "@adrieni : Your same implementation can be found from This Notebook series
\nRefer this Discussion if you are confused on \"How to make a submission with hidden folders\"
", "rawMarkdown": "adrieni : Your same implementation can be found from [This Notebook series](https://www.kaggle.com/timothyalexjohn/birdsong-classifier-keras-cnn-part-1-notebook-1)\n\nRefer this[ Discussion](https://www.kaggle.com/c/birdsong-recognition/discussion/175169#975224) if you are confused on \"How to make a submission with hidden folders\"", "votes": null } ], "comments": [ { "id": 975238, "author_name": "timothyalexjohn", "author_url": "", "post_date": "08/18/2020 08:00:57", "content": "@adrieni : Your same implementation can be found from This Notebook series
\nRefer this Discussion if you are confused on \"How to make a submission with hidden folders\"
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "901396": "Hello everyone,\nI have a little problem. Indeed, I don't know how I can test on the test dataset, because I didn't understand how to use the two audio files. My model uses spectrograms, so I take each audio extract and transform it into a spectrogram. \nHowever, I don't understand how to do this on the two audio files. Should I divide them into small clips and then turn them into a spectrogram and then predict which bird sings in each clip, or should I do something else?\n\nPS: Should I download a test dataset? (I saw on a notebook that it had a dataset with lots of little clips, is that what I'm looking for)?", "975238": "adrieni : Your same implementation can be found from [This Notebook series](https://www.kaggle.com/timothyalexjohn/birdsong-classifier-keras-cnn-part-1-notebook-1)\n\nRefer this[ Discussion](https://www.kaggle.com/c/birdsong-recognition/discussion/175169#975224) if you are confused on \"How to make a submission with hidden folders\"" }, "source": "meta" }