Hi, i was using HTK for gait recognition and i having some problem with hvite and herest.

Below was the step that i have go through:

1)hcompv -m model/proto/hmmfile model/d11-1x.csv.bin model/d11-2x.csv.bin model/d11-3x.csv.bin

where hmm file was the protype definition
where d11-1x,2x,3x.csv.bin was the data to train it

Below was the prototype definition i used:

<BeginHMM>
<VecSize> 26 <USER>
<NumStates> 6
<State> 2
<Mean> 26
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
<Variance> 26
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
<State> 3
<Mean> 26
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
<Variance> 26
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
<State> 4
<Mean> 26
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
<Variance> 26
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
<State> 5
<Mean> 26
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
<Variance> 26
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
<TransP> 6
0.0 0.5 0.5 0.0 0.0 0.0
0.0 0.4 0.3 0.3 0.0 0.0
0.0 0.0 0.4 0.3 0.3 0.0
0.0 0.0 0.0 0.4 0.3 0.3
0.0 0.0 0.0 0.0 0.5 0.5
0.0 0.0 0.0 0.0 0.0 0.0
<EndHMM>


2)Hrest hmmfile model/d11-4x.csv.bin model/d11-5x.csv.bin model/d11-7x.csv.bin

Then i generate the dictionary and network as:

$GROUP = ENROLLEE | RESTOFALL;
( $GROUP )

(gram.txt)

3)hparse model/network/gram.txt model/network/net.slf

Which generate the network named "net.slf"

And below was the dictionary file content:

ENROLLEE [ENROLLEE] hmmfile
RESTOFALL [RESTOFALL] hmmfiletwo


(dict.txt)
(Among all the model, i just need to identify whether they are categorize as enrollee or restofall group. )

###Then the problem comes here, i suppose use hvite to start the analysis but i keep getting errors

4)hvite model/network/net.slf model/dictionary/dict.txt model/hmmlist/hmmlist (What is missing here)

-model/network/net.slf was the network that i have generated

-model/dictionary/dict.txt was the dictionary

-model/hmmlist/hmmlist was the hmm model list and  the content was like :

C:/Users/Obana/Downloads/HTK-3.4.1/htk/bin.win32/model/hmm0/hmmfile
C:/Users/Obana/Downloads/HTK-3.4.1/htk/bin.win32/model/hmm0/hmmfile2

Output of the command above was:

error 7060: Expected new line after 1th Hmm

error 2321

###I have read most of the tutorial online but most of them are based on speech recognition...Im new to HTK and please help me out if you know how to do it.. I was learning HTK for weeks and only manage to generate the HMM model and now i just want to start the recognition as fast as possible.

###Please let me know any necessary step that i have miss and describe it to me as details as possible since i was new to HTK. I couldnt find anyone to help me out ......so please help me if you know how to do :(

--- (Edited on 12/5/2011 8:07 am [GMT-0600] by obana1990) ---

--- (Edited on 12/5/2011 8:14 am [GMT-0600] by obana1990) ---