[Olympus developers 107]: Re: description of features

[Thomas Harris]

Some of these, the ones that Helios actually produces, are described at
http://wiki.speech.cs.cmu.edu/olympus/index.php/Helios

Many of these however, although logged by Helios, are features taken from
other Olympus components like the Audio Server and Parser. There are
placeholders for the Audio Server and Phoenix feature documentation at the
following locations
http://wiki.speech.cs.cmu.edu/olympus/index.php/Final_ASR_result_frame
http://wiki.speech.cs.cmu.edu/olympus/index.php/Partial_ASR_result_frame
http://wiki.speech.cs.cmu.edu/olympus/index.php/Parsed_final_ASR_result_frame
http://wiki.speech.cs.cmu.edu/olympus/index.php/Parsed_partial_ASR_result_frame
but no actual documentation yet.

Thanks,
Thomas

On Thu, Apr 2, 2009 at 10:45 AM, Tiziana Ligorio
<ligoriotiziana at gmail.com>wrote:

> Hello,does anyone have a pointer to a full description of some of the
> features found in helios.log?
> Specifically I am looking for a definition, the range of possible values
> and the type of the field (precision, integer, sign) of the following
> features:
>
>  am_score   avg_wordconf   avg_validwordconf
>  confidence   decoder_score   frame_num   garble_clear   h_avg_confidence
> h_avg_gap_num   lm_score   max_wordconf   min_wordconf   speak_rate
>  speak_rate_phones   speak_rate_syl   total_num_parses   word_num
> words_per_slot
>
>
>
> Thank you,
> Tiziana Ligorio
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.srv.cs.cmu.edu/pipermail/olympus-developers/attachments/20090402/c300ecdb/attachment.html