Devanagari Characters speech pronunciation dataset. This dataset was collected from Teach Me Nepali webapp. Go to the link to contribute your pronunciation/voice recording to the dataset.
This dataset contains 14,452 data samples corresponding to 444 classes. The labels mapping to the actual characters is in the label_to_character.csv file.
### dictionary of characters
dict_1 = { 'a':'अ', 'aa':'आ', 'i':'इ', 'ii':'ई', 'u':'उ', 'uu':'ऊ', 'e':'ए', 'ai':'ऐ', 'o':'ओ', 'au':'औ', 'am':'अं', 'ah':'अः' }
dict_2 = { 'aa':'ा', 'i':'ि', 'ii':'ी', 'u':'ु', 'uu':'ू', 'e':'े', 'ai':'ै', 'o':'ो', 'au':'ौ', 'am':'ं', 'ah':'ः' }
dict_3 = { 'ka':'क', 'kha':'ख', 'ga':'ग', 'gha':'घ', 'nga':'ङ',
'cha':'च', 'chha':'छ', 'ja':'ज', 'jha':'झ', 'yan':'ञ',
'toa':'ट', 'thoa':'ठ', 'doa':'ड', 'dhoa':'ढ', 'ndoa':'ण',
'ta':'त', 'tha':'थ', 'da':'द', 'dha':'ध', 'na':'न',
'pa':'प', 'fa':'फ', 'ba':'ब', 'bha':'भ', 'ma':'म',
'ya':'य', 'ra':'र', 'la':'ल', 'wa':'व',
'sa':'ष', 'sha':'स', 'shaa':'श', 'ha':'ह',
'chya':'क्ष', 'tra':'त्र', 'gya':'ज्ञ' }
Thanks to all those who helped to collect this dataset. Without their contribution this dataset was not possible. CREDITS