-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathREADME
112 lines (93 loc) · 3.02 KB
/
README
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
$Id: README,v 1.5 1998/03/16 08:21:26 kanungo Exp kanungo $
Package: UMDHMM version 1.02
Author: Tapas Kanungo (kanungo@cfar.umd.edu)
Organization: University of Maryland, Collge Park, MD
Web: http://www.cfar.umd.edu/~kanungo
Date: 19 February, 1998
Updated on 5 May, 1999: see CHANGES file.
Updated on 6 May, 1999: see CHANGES file.
This software contains code for understanding the basics
of hidden Markov models (HMM). The notation used is
very similar to that used by Rabiner and Juang in:
- Rabiner, L. R. and B. H. Juang, "Fundamentals of Speech Recognition,"
Prentice Hall, 1993.
- Rabiner, L. R., "A Tutorial on Hidden Markov Models and Selected
Applications in Speech Recognition, Prov. of IEEE, vol. 77, no. 2,
pp. 257-286, 1989.
- Rabiner, L. R., and B. H. Juang, "An Introduction to Hidden Markov Models,"
IEEE ASSP Magazine, vol. 3, no. 1, pp. 4-16, Jan. 1986.
---------------------------------------------
Installation:
---------------------------------------------
--------------------
UNIX: Dec, Sun Solaris, Linux (redhat):
--------------------
Type "make all" at the unix prompt. It should
compile the package.
--------------------
Microsoft NT/95/98:
--------------------
1. Get the GNU package from:
ftp://go.cygnus.com/pub/sourceware.cygnus.com/cygwin/latest/full.exe
This package includes gcc and various commands and
shells (sh, bash, etc.) that make the PC have a unix
like environment.
2. Change to the UMDHMM directory and type "make all".
---------------------------------------------
Executables:
---------------------------------------------
genseq: Generates a symbol sequence using the specified model
testvit: Generates the most like state sequence for a given symbol sequence,
given the HMM, using Viterbi.
esthmm: Estimates the HMM from a given symbol sequence using BaumWelch.
testfor: Computes log Prob(observation|model) using the Forward algorithm.
Note 1: The model test.hmm and sequence test.seq solve exercise 6.3 in
the book by Rabiner and Juang (page 341). Just execute the command:
prompt% testvit test.hmm test.seq
and compare the output with the solution given in the book.
---------------------------------------------
HMM file format:
---------------------------------------------
M= <number of symbols>
N= <number of states>
A:
a11 a12 ... a1N
a21 a22 ... a2N
. . . .
. . . .
. . . .
aN1 aN2 ... aNN
B:
b11 b12 ... b1M
b21 b22 ... b2M
. . . .
. . . .
. . . .
bN1 bN2 ... bNM
pi:
pi1 pi2 ... piN
---------------------------------------------
Sample HMM file:
---------------------------------------------
M= 2
N= 3
A:
0.333 0.333 0.333
0.333 0.333 0.333
0.333 0.333 0.333
B:
0.5 0.5
0.75 0.25
0.25 0.75
pi:
0.333 0.333 0.333
---------------------------------------------
Sequence file format:
---------------------------------------------
T=<seqence lenght>
o1 o2 o3 . . . oT
---------------------------------------------
Sample sequence file:
---------------------------------------------
T= 10
1 1 1 1 2 1 2 2 2 2