Dataset Viewer
Auto-converted to Parquet Duplicate
name
stringlengths
14
28
speaker_embeddings
sequencelengths
512
512
ARA_NORM__0001
[ -0.0659799799323082, 0.04209241271018982, 0.03229561448097229, 0.04620331898331642, 0.01768263429403305, -0.024051379412412643, -0.047923509031534195, -0.003332102205604315, 0.037848394364118576, 0.016594087705016136, -0.07967960089445114, -0.07822584360837936, 0.04748737812042236, 0.01669...
ARA_NORM__0002
[ -0.06973488628864288, 0.0659625232219696, 0.027099382132291794, 0.04050631821155548, 0.006591961719095707, -0.023426560685038567, -0.04586838558316231, -0.00008750687266001478, 0.023234965279698372, 0.015776561573147774, -0.08305037021636963, -0.08252853900194168, 0.04135022312402725, 0.01...
ARA_NORM__0003
[ -0.07133269309997559, 0.07348307967185974, 0.02331993356347084, 0.03933476656675339, 0.004754405003041029, -0.015827365219593048, -0.04054034873843193, 0.012947700917720795, 0.03335432708263397, 0.015283261425793171, -0.0818214863538742, -0.08160760998725891, 0.04470960050821304, 0.0148251...
ARA_NORM__0004
[ -0.06734980642795563, 0.045863233506679535, -0.0037403274327516556, 0.037173669785261154, -0.008371586911380291, -0.017284179106354713, -0.040143318474292755, -0.012732114642858505, 0.01392578985542059, 0.015175282023847103, -0.0805085301399231, -0.08132199943065643, 0.03924332931637764, 0...
ARA_NORM__0005
[ -0.06971703469753265, 0.05831428989768028, 0.0151000851765275, 0.04346131160855293, -0.003935683984309435, -0.023676451295614243, -0.04661397263407707, 0.006905467715114355, 0.023445608094334602, 0.014888489618897438, -0.08714721351861954, -0.09585351496934891, 0.037442147731781006, 0.0116...
ARA_NORM__0006
[ -0.07249975204467773, 0.06346861273050308, 0.020976845175027847, 0.050916366279125214, -0.018440866842865944, -0.01464331429451704, -0.04113798961043358, 0.007510109804570675, 0.030774567276239395, 0.014803653582930565, -0.08497746288776398, -0.08081210404634476, 0.0442005880177021, 0.0101...
ARA_NORM__0007
[ -0.06843354552984238, 0.06229601800441742, 0.018756048753857613, 0.037696558982133865, 0.014885245822370052, -0.019326603040099144, -0.043232545256614685, 0.005685409065335989, 0.018815483897924423, 0.01491484697908163, -0.08353065699338913, -0.08706170320510864, 0.04658128321170807, 0.021...
ARA_NORM__0008
[ -0.068301260471344, 0.0674300417304039, 0.02195860631763935, 0.04290853068232536, 0.005887266248464584, -0.021220073103904724, -0.04337047040462494, 0.006714160088449717, 0.031993720680475235, 0.015572967939078808, -0.07650937885046005, -0.08130848407745361, 0.044800546020269394, 0.0140240...
ARA_NORM__0009
[ -0.07151755690574646, 0.06436961144208908, 0.029972214251756668, 0.04101802408695221, 0.0010736663825809956, -0.02372523583471775, -0.04044124484062195, 0.004492100328207016, 0.03974296897649765, 0.01493423618376255, -0.07434023916721344, -0.07406779378652573, 0.041173305362463, 0.01003202...
ARA_NORM__0010
[ -0.0660473108291626, 0.06662487238645554, 0.007035648450255394, 0.044822871685028076, 0.0098688043653965, -0.020464612171053886, -0.04262413829565048, 0.003334800014272332, 0.014395471662282944, 0.015439573675394058, -0.08254951238632202, -0.08512959629297256, 0.050061389803886414, 0.02190...
ARA_NORM__0011
[ -0.0650608241558075, 0.07053963094949722, 0.022372892126441002, 0.04088225215673447, -0.00013727931946050376, -0.02965785190463066, -0.03798980265855789, 0.016222428530454636, 0.031387362629175186, 0.015148621052503586, -0.08551652729511261, -0.09186308830976486, 0.04188187047839165, 0.016...
ARA_NORM__0012
[ -0.0662936270236969, 0.05959491804242134, 0.01726270094513893, 0.03889947012066841, 0.0012256171321496367, -0.02486160583794117, -0.031091364100575447, 0.01414320059120655, 0.026143057271838188, 0.016140636056661606, -0.08013118803501129, -0.07948856800794601, 0.04097095504403114, 0.010271...
ARA_NORM__0013
[ -0.07570622116327286, 0.0681443139910698, 0.013206860050559044, 0.03901365026831627, -0.004729507025331259, -0.018175041303038597, -0.04645208269357681, 0.0068206689320504665, 0.03348199650645256, 0.015534907579421997, -0.08156989514827728, -0.08884923160076141, 0.04241437837481499, 0.0070...
ARA_NORM__0014
[ -0.05657226964831352, 0.059136684983968735, 0.009539851918816566, 0.03195657581090927, -0.005536873824894428, -0.024160509929060936, -0.04141028970479965, 0.021783407777547836, 0.02218497171998024, 0.01605508290231228, -0.08130393177270889, -0.07357858866453171, 0.04905012249946594, 0.0161...
ARA_NORM__0015
[ -0.07760477811098099, 0.06168088689446449, 0.009273986332118511, 0.030884141102433205, 0.005122964736074209, -0.027855481952428818, -0.035848721861839294, 0.011220757849514484, 0.03296290710568428, 0.01622156985104084, -0.07275546342134476, -0.08268661051988602, 0.04266272112727165, 0.0172...
ARA_NORM__0016
[ -0.06942803412675858, 0.07019992917776108, 0.015379711985588074, 0.03616848215460777, 0.010792330838739872, -0.02831064909696579, -0.04308127239346504, 0.010385483503341675, 0.034438736736774445, 0.0162182804197073, -0.07996232807636261, -0.08284735679626465, 0.043309785425662994, 0.007371...
ARA_NORM__0017
[ -0.0675847977399826, 0.0669979676604271, 0.01937234401702881, 0.04274537041783333, -0.004619080573320389, -0.0261896550655365, -0.043008629232645035, 0.008960394188761711, 0.028711190447211266, 0.01506789866834879, -0.07975904643535614, -0.08736709505319595, 0.04539545997977257, 0.01309748...
ARA_NORM__0018
[ -0.06410190463066101, 0.06589525938034058, 0.020638924092054367, 0.026094675064086914, 0.00004839018947677687, -0.017271120101213455, -0.04181089624762535, 0.008957061916589737, 0.029986942186951637, 0.013658421114087105, -0.07322414219379425, -0.09212219715118408, 0.04194618761539459, 0.0...
End of preview. Expand in Data Studio

Arabic Speaker Embeddings extracted from ASC and ClArTTS

There is one speaker embedding for each utterance in the validation set of both datasets. The speaker embeddings are 512-element X-vectors.

Arabic Speech Corpus has 100 files for a single male speaker and ClArTTS has 205 files for a single male speaker.

The X-vectors were extracted using this script, which uses the speechbrain/spkrec-xvect-voxceleb model.

Usage:

from datasets import load_dataset

embeddings_dataset = load_dataset("herwoww/arabic_xvector_embeddings", split="validation")
speaker_embedding = torch.tensor(embeddings_dataset[1]["speaker_embeddings"]).unsqueeze(0)
Downloads last month
1,337

Spaces using herwoww/arabic_xvector_embeddings 4