Search for tag: "speech recognition"

15 Media

ASR Lecture 14 duration 54 minutes 41 seconds ASR Lecture 14 ASR Lecture 14	ASR Lecture 14 ASR Lecture 14: Sequence discriminative training speech recognition sequence model probability word likelihood training estimation function hmm lattice mmi statistics topology denominator probabilities network arc bit level lattices From Peter Bell 0 likes 103 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created March 4th, 2021
ASR Lecture 13 duration 58 minutes 6 seconds ASR Lecture 13 ASR Lecture 13	ASR Lecture 13 ASR Lecture 13: Speaker adaptation speech recognition speaker adaptation model parameters transform speech vector network methods gaussians term training idea linear scheme space bit gaussian variance matrix From Peter Bell 0 likes 82 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created March 29th, 2021
ASR Lecture 12 duration 55 minutes 49 seconds ASR Lecture 12 ASR Lecture 12	ASR Lecture 12 ASR Lecture 12: Neural Networks for Acoustic Modelling 3 speech recognition features network weights layer units unit context gate networks input model output sort point frame kind bit frames lecture lstms From Peter Bell 0 likes 88 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created February 25th, 2021
ASR Lecture 10 (part 2); Lecture 11 duration 57 minutes 0 seconds ASR Lecture 10 (part 2); Lecture 11 ASR Lecture 10 (part 2); Lecture 11	ASR Lecture 10 (part 2); Lecture 11 ASR Lecture 10 (part 2): Neural network models (1); ASR Lecture 11: Neural network models (2) speech recognition network networks probability bit speech recognition features output layer error weights outputs label delta things function hmm term lecture layers From Peter Bell 0 likes 106 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created February 22nd, 2021
ASR Lecture 8 duration 51 minutes 43 seconds ASR Lecture 8 ASR Lecture 8	ASR Lecture 8 ASR Lecture 8: Large vocabulary ASR speech recognition word words model path viterbi language probability pin bit algorithm end speech recognition vocabulary thing transitions probabilities step models decoding From Peter Bell 0 likes 149 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created February 4th, 2021
ASR Lecture 7 duration 56 minutes 24 seconds ASR Lecture 7 ASR Lecture 7	ASR Lecture 7 ASR Lecture 7: Context-dependent phone models speech recognition From Peter Bell 0 likes 108 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created February 1st, 2021
ASR Lecture 9: missing content duration 6 minutes 19 seconds ASR Lecture 9: missing content ASR Lecture 9: missing content	ASR Lecture 9: missing content Missing content from ASR Lecture 9 on back-off bigram LMs speech recognition backoff word bigram vocabulary probability history weight language pizza probabilities model factor property array hash arc peck peter arts poverty From Peter Bell 0 likes 94 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created February 15th, 2021
ASR Lecture 9 part II; ASR Lecture 10 part I duration 52 minutes 59 seconds ASR Lecture 9 part II; ASR Lecture 10 part I ASR Lecture 9 part II; ASR Lecture 10 part I	ASR Lecture 9 part II; ASR Lecture 10 part I Second part of ASR Lecture 9: WFSTs for ASR First part of ASR Lecture 10: Neural network acoustic models 1 speech recognition error context function bit vector output weight score point thing phones weights grandma respect t. training things hmm space pizza From Peter Bell 0 likes 193 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created February 11th, 2021
ASR Lecture 9 part I duration 47 minutes 46 seconds ASR Lecture 9 part I ASR Lecture 9 part I	ASR Lecture 9 part I ASR Lecture 9: WFSTs for ASR (partial lecture due to technical issues) speech recognition transducer probability bit output symbol weights probabilities point composition word path log input sequence weight pizza outputs grandma recognition loop From Peter Bell 0 likes 185 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created February 8th, 2021
ASR Lecture 6 duration 1 hour 2 minutes ASR Lecture 6 ASR Lecture 6	ASR Lecture 6 ASR Lecture 6: Gaussian Mixture Models speech recognition From Peter Bell 0 likes 125 plays 0 Licence Type Creative Commons - Attribution Non Commercial
ASR Lecture 4 duration 55 minutes 42 seconds ASR Lecture 4 ASR Lecture 4	ASR Lecture 4 ASR lecture 4: Introduction to HMMs speech recognition model probability observation topology sequence step observations hmm bit t. paths algorithm speech projects probabilities recognition words question likelihood word From Peter Bell 0 likes 162 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created January 21st, 2021
ASR Lecture 3 duration 1 hour 32 minutes ASR Lecture 3 ASR Lecture 3	ASR Lecture 3 ASR Lecture 3: Speech signal analysis (2) speech recognition features window spectrum power frequency analysis asr windows mfcc frequencies delta output questions dft linear scale bit mfccs difference question From Peter Bell 0 likes 129 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created January 18th, 2021
ASR Lecture 5 duration 55 minutes 33 seconds ASR Lecture 5 ASR Lecture 5	ASR Lecture 5 ASR Lecture 5: HMM algorithms speech recognition From Peter Bell 0 likes 164 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created January 25th, 2021
ASR Lecture 2 duration 56 minutes 58 seconds ASR Lecture 2 ASR Lecture 2	ASR Lecture 2 ASR Lecture 2: Speech signal analysis (1) speech recognition speech frequency frequencies signals peaks characteristics sound signal impulse power analysis spectrum features process philtres philtre axis kilohertz question train From Peter Bell 0 likes 215 plays 0 Licence Type Creative Commons - Attribution Non Commercial Date Created January 14th, 2021
Deep Fakes in the age of Voice-Enabled Consumer… duration 10 seconds Deep Fakes in the age of Voice-Enabled Consumer… Deep Fakes in the age of Voice-Enabled Consumer Electronics - Jennifer Williams	Deep Fakes in the age of Voice-Enabled Consumer Electronics - Jennifer Williams Controversies in the Data Society 2020 Week 3 voice synthesis speech recognition speech synthesis cybercrime spoofing evidence From James Stewart 0 likes 22 plays 0 Publisher UoE Licence Type Creative Commons - Attribution Non Commercial No Derivatives Language English Date Created January 31st, 2020