next up previous contents index
Next: 16.3 Syntax Up: 16 HTK Standard Lattice Format (SLF) Previous: 16.1 SLF Files

16.2 Format

 

The format  is designed to allow optional information that at its most detailed gives full identity, alignment and score (log likelihood) information at the word and phone level to allow calculation of the alignment and likelihood of an individual hypothesis. However, without scores or times the lattice is just a word graph. The format is designed to be extendible. Further field names can be defined to allow arbitrary information to be added to the lattice without making the resulting file unreadable by others.

The lattices are stored in a text file as a series of fields that form two blocks:

Either block may contain comment lines , for which the first character is a `#' and the rest of the line is ignored.

All non-comment lines consist of fields, separated by white space. Fields consist of an alphanumeric field name, followed by a delimiter (the character `=' or ` ') and a (possibly ``quoted'') field value. All letters in the name are significant. Single character field names are reserved for fields defined in the specification and single character abbreviations may be used for many of the fields defined below.

The convention used to define the current field names  is that lower case is used for optional fields and upper case is used for required fields. The meaning of field names can be dependent on the context in which they appear.

The header must include a field specifying which utterance was used to generate the lattice and a field specifying the version of the lattice specification used. The header is terminated by a line which defines the number of nodes and links in the lattice.

The node definitions are optional but if included each node definition consists of a single line which specifies the node number followed by optional fields that may (for instance) define the time of the node or the word hypothesis ending at that node.

The link definitions are required and each link definition consists of a single line which specifies the link number as well as the start and end node numbers that it connects to and optionally other information about the link such as the word identity and language model score. If word identity information is not present in node definitions then it must appear in link definitions.


next up previous contents index
Next: 16.3 Syntax Up: 16 HTK Standard Lattice Format (SLF) Previous: 16.1 SLF Files

ECRL HTK_V2.1: email [email protected]