srvk
word Lattice to phone lattice / to slf
Hi Alls, I would like to convert a word lattice generated using eesen to phone level lattice like in kaldi (lattice-to-phone-lattice 1.mdl ark:1.lats ark:phones.lats). Is there a sollution? Is it possible to convert a lattice also generated by eesen to slf format (like this generated by htk)? Thank you in advance.
Apologies for the late response, I'd been traveling a lot. I think it should be possible to generate a "fake" model file that corresponds to the Eesen context independent acoustic model and which would allow you to run the Kaldi tools for that process. Did you try that?
Thank you Mr Florian for your response. A I will try this.
2017-03-23 4:04 GMT+01:00 Florian Metze [email protected]:
Apologies for the late response, I'd been traveling a lot. I think it should be possible to generate a "fake" model file that corresponds to the Eesen context independent acoustic model and which would allow you to run the Kaldi tools for that process. Did you try that?
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/srvk/eesen/issues/125#issuecomment-288604166, or mute the thread https://github.com/notifications/unsubscribe-auth/AO5KDCFPDDRvIDD5B_CuVs4AIkzYt051ks5roeE2gaJpZM4MUl8U .
-- Best regards, Sana Khamekhem Jemni
PhD Student MIRACL Laboratory ENIS
Hi, I have generated a model using HMM(kaldi), but I have this issue when calling Kaldi functions:
lattice-align-words-lexicon --max-expand=10.0 data/lang_phn_test_3/phones/align_lexicon.int exp/mono0a/final.mdl ark:- ark:- WARNING (lattice-align-words-lexicon[5.1]:OutputPhoneArc():phone-align-lattice.cc:310) 44 -> 1 WARNING (lattice-align-words-lexicon[5.1]:OutputPhoneArc():phone-align-lattice.cc:311) Phone changed before final transition-id found [broken lattice or mismatched model or wrong --reorder option?] ASSERTION_FAILED (lattice-align-words-lexicon[5.1]:Advance():word-align-lattice-lexicon.cc:581) : 'phone == tmodel.TransitionIdToPhone(tids.back()) && "Error: lattice is not phone-aligned."'
Maybe something to do with the three-state vs single-state topology? I must admit I am not familiar with these Kaldi tools - sorry