search

MiteshPuthran / Speech-Emotion-Analyzer

The neural network model is capable of detecting five different male/female emotions from audio speeches. (Deep Learning, NLP, Python)
MIT License
1.29k stars 436 forks source link

Having doubt regarding the Rawdata in code :final_results_gender_test.ipynb #1

Closed vikash512 closed 6 years ago

vikash512 commented 6 years ago

having doubt regarding these lines, what is the data in that and which format .

mylist= os.listdir('RawData/')

data, sampling_rate = librosa.load('RawData/f11 (2).wav')

MiteshPuthran commented 6 years ago

Hello vikash512,

The data is audio files in .wav format. The data source has been linked in the readme file. In the second line the librosa library helps to extract data from the audio files, which is stored in as time in second and its sampling rate.

Arunpoochelvan commented 5 years ago

Hi, Can you please share us the RawData/ file? as we don't see 1800 audio files when combining both the dataset you mentioned on readme

vlkare commented 5 years ago

Yes please do share the file

ejlok1 commented 5 years ago

The raw SAVEE dataset is seperated into 4 folders namely DC, JE, KJ, KL. What Mitesh has done here is combine them all into a single folder. Because there's a clash in the naming of the files, The audio file f10.wav for instance exist in all 4 folders. hence why the file when combine goes f10 (2).wav

vlkare commented 5 years ago

Means I have to combine all the files from the 4 folder into one?? And where is RAVDESS data used??

On Mon, 2 Sep 2019, 04:23 Eu Jin Lok, notifications@github.com wrote:

The raw SAVEE dataset is seperated into 4 folders namely DC, JE, KJ, KL. What Mitesh has done here is combine them all into a single folder. Because there's a clash in the naming of the files, The audio file f10.wav for instance exist in all 4 folders. hence why the file when combine goes f10 (2).wav

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/MITESHPUTHRANNEU/Speech-Emotion-Analyzer/issues/1?email_source=notifications&email_token=AJS6YNJMME3RWIGKLZ2SZCTQHRBV7A5CNFSM4FN35ZLKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5UMMNY#issuecomment-526960183, or mute the thread https://github.com/notifications/unsubscribe-auth/AJS6YNIVBLMV5Q46UE42AC3QHRBV7ANCNFSM4FN35ZLA .

vlkare commented 5 years ago

RAVDESS data and SAVEE data has to combine in a same file.??? Thats is what it is in RawData/???

On Mon, 2 Sep 2019, 09:20 Vikram Kare, vlkare@mitaoe.ac.in wrote:

Means I have to combine all the files from the 4 folder into one?? And where is RAVDESS data used??

On Mon, 2 Sep 2019, 04:23 Eu Jin Lok, notifications@github.com wrote:

The raw SAVEE dataset is seperated into 4 folders namely DC, JE, KJ, KL. What Mitesh has done here is combine them all into a single folder. Because there's a clash in the naming of the files, The audio file f10.wav for instance exist in all 4 folders. hence why the file when combine goes f10 (2).wav

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/MITESHPUTHRANNEU/Speech-Emotion-Analyzer/issues/1?email_source=notifications&email_token=AJS6YNJMME3RWIGKLZ2SZCTQHRBV7A5CNFSM4FN35ZLKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5UMMNY#issuecomment-526960183, or mute the thread https://github.com/notifications/unsubscribe-auth/AJS6YNIVBLMV5Q46UE42AC3QHRBV7ANCNFSM4FN35ZLA .

ejlok1 commented 5 years ago

Yes correct they are all in the same folder. The for loop iterates through various if statements. It first goes through the RAVDESS file format first, then SAVEE as the last bunch.