.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.36.
.TH AUDIODB "1" "September 2008" "audioDB version 0.8.preview" "User Commands"
.SH NAME
audioDB \- manual page for audioDB version 0.8.preview
.SH SYNOPSIS
.B audioDB
[\fIOPTIONS\fR]...
.SH DESCRIPTION
audioDB version 0.8.preview
.PP
A feature vector database management system for content\-based retrieval.
.TP
\fB\-\-full\-help\fR
Print help, including hidden options, and exit
.TP
\fB\-V\fR, \fB\-\-version\fR
Print version and exit
.TP
\fB\-H\fR, \fB\-\-help\fR
print help on audioDB usage and exit.
.TP
\fB\-v\fR, \fB\-\-verbosity\fR=\fIdetail\fR
level of detail of operational information.
(default=`1')
.SS "Database Operations:"
.IP
All database operations require a database argument.
.IP
Database commands are UPPER CASE. Command options are lower case.
.TP
\fB\-d\fR, \fB\-\-database\fR=\fIfilename\fR
database file required by Database commands.
.TP
\fB\-\-adb_root\fR=\fIpath\fR
path prefix for database
.SS "Database Creation:"
.IP
Creating a new database file.
\fB\-N\fR, \fB\-\-NEW\fR                     make a new (initially empty) database.
.TP
\fB\-\-datasize\fR=\fIINT\fR
size of data table requested (in MB)
(default=`1355')
.TP
\fB\-\-ntracks\fR=\fIINT\fR
capacity of database for tracks
(default=`20000')
.TP
\fB\-\-datadim\fR=\fIINT\fR
dimensionality of stored data  (default=`9')
.SS "Database Maintenance:"
.IP
Tweaking and dumping databases.
\fB\-D\fR, \fB\-\-DUMP\fR                    output all entries: index key size.
.TP
\fB\-\-output\fR=\fISTRING\fR
output directory  (default=`audioDB.dump')
.TP
\fB\-L\fR, \fB\-\-L2NORM\fR
unit norm vectors and norm all future inserts.
.TP
\fB\-P\fR, \fB\-\-POWER\fR
turn on power flag for database.
.TP
\fB\-Z\fR, \fB\-\-LISZT\fR
LIst keyS and siZes of Tracks
.TP
\fB\-\-lisztOffset\fR=\fInumber\fR
LISZT track offset (0\-based index)
(default=`0')
.TP
\fB\-\-lisztLength\fR=\fInumber\fR
number of LISZT items to return  (default=`32')
.SS "Database Information:"
.IP
Information about databases.
\fB\-S\fR, \fB\-\-STATUS\fR                  output database information to stdout.
.TP
\fB\-\-SAMPLE\fR
sample statistics for database.
.TP
\fB\-\-nsamples\fR=\fInumber\fR
number of pairwise samples to take.
(default=`2000')
.SS "Database Insertion:"
.IP
The following commands insert feature files, with optional keys and
timestamps.
.TP
\fB\-I\fR, \fB\-\-INSERT\fR
add feature vectors to an existing database.
.TP
\fB\-\-adb_feature_root\fR=\fIpath\fR
path prefix for feature files, times files and
power files
.TP
\fB\-f\fR, \fB\-\-features\fR=\fIfilename\fR
binary series of vectors file {int sz:ieee
double[][sz]:eof}.
.TP
\fB\-t\fR, \fB\-\-times\fR=\fIfilename\fR
list of time points (ascii) for feature
vectors.
.TP
\fB\-w\fR, \fB\-\-power\fR=\fIfilename\fR
binary power feature file.
.TP
\fB\-k\fR, \fB\-\-key\fR=\fIidentifier\fR
unique identifier associated with features.
.TP
\fB\-B\fR, \fB\-\-BATCHINSERT\fR
add feature vectors named in a \fB\-\-featureList\fR
file (with optional keys in a \fB\-\-keyList\fR file)
to the named database.
.TP
\fB\-F\fR, \fB\-\-featureList\fR=\fIfilename\fR
text file containing list of binary feature
vector files to process, one per track
.TP
\fB\-T\fR, \fB\-\-timesList\fR=\fIfilename\fR
text file containing list of ascii \fB\-\-times\fR for
each \fB\-\-features\fR file in \fB\-\-featureList\fR.
.TP
\fB\-W\fR, \fB\-\-powerList\fR=\fIfilename\fR
text file containing list of binary power
feature file.
.TP
\fB\-K\fR, \fB\-\-keyList\fR=\fIfilename\fR
text file containing list of unique identifiers
associated with \fB\-\-features\fR.
.SS "Database Search:"
.IP
These commands control the retrieval behaviour.
.TP
\fB\-Q\fR, \fB\-\-QUERY\fR=\fIsearchtype\fR
content\-based search on \fB\-\-database\fR using
\fB\-\-features\fR as a query. Optionally restrict
the search to those tracks identified in a
\fB\-\-keyList\fR.  (possible values="point",
"track", "sequence", "nsequence",
"onetoonensequence")
.TP
\fB\-p\fR, \fB\-\-qpoint\fR=\fIposition\fR
ordinal position of query start point in
\fB\-\-features\fR file.  (default=`0')
.TP
\fB\-n\fR, \fB\-\-pointnn\fR=\fInumpoints\fR
number of point nearest neighbours to use in
retrieval.  (default=`10')
.TP
\fB\-R\fR, \fB\-\-radius\fR=\fIDOUBLE\fR
radius search, returns all
points/tracks/sequences inside given radius.
(Overrides \fB\-\-pointnn\fR).  (default=`1.0')
.TP
\fB\-r\fR, \fB\-\-resultlength\fR=\fIlength\fR
maximum length of the result list.
(default=`10')
.TP
\fB\-l\fR, \fB\-\-sequencelength\fR=\fIlength\fR
length of sequences for sequence search.
(default=`16')
.TP
\fB\-h\fR, \fB\-\-sequencehop\fR=\fIhop\fR
hop size of sequence window for sequence
search.  (default=`1')
.TP
\fB\-\-absolute\-threshold\fR=\fIDOUBLE\fR
absolute power threshold for consideration of
.IP
query or target sequence (in Bels)
.TP
\fB\-\-relative\-threshold\fR=\fIDOUBLE\fR
relative power threshold between query and
.IP
target sequence (in Bels)
.SS "Locality-sensitive hashing (LSH) parameters:"
.IP
These parameters control LSH indexing and retrieval
.TP
\fB\-X\fR, \fB\-\-INDEX\fR
build an index for \fB\-d\fR database at \fB\-R\fR radius and
\fB\-l\fR sequenceLength
.TP
\fB\-\-lsh_k\fR=\fIsize\fR
even number of independent hash functions to
employ with LSH  (default=`8')
.TP
\fB\-\-lsh_m\fR=\fIsize\fR
number of hash tables is m(m\-1)/2
(default=`5')
.TP
\fB\-\-lsh_N\fR=\fIsize\fR
number of rows per hash tables
(default=`100000')
.TP
\fB\-\-lsh_b\fR=\fIsize\fR
number of tracks per indexing iteration
(default=`500')
.TP
\fB\-\-lsh_exact\fR
use exact evaluation of points retrieved by
LSH.  (default=off)
.TP
\fB\-\-lsh_on_disk\fR
Construct LSH hash tables for on\-disk query
(INDEX/QUERY)  (default=off)
.TP
\fB\-\-lsh_use_u_functions\fR
use m independent hash functions
combinatorically to approximate L independent
hash functions.  (default=off)
.SS "Normalization control parameters:"
.IP
These parameters control the normalization of feaures at query time
.TP
\fB\-\-no_unit_norming\fR
do not unit norm features when querying an
L2Norm databases.  (default=off)
.SS "Web Services:"
.IP
These commands enable the database process to establish a connection via the
internet and operate as separate client and server processes.
.TP
\fB\-s\fR, \fB\-\-SERVER\fR=\fIport\fR
run as standalone web service on named port.
(default=`14475')
.TP
\fB\-\-load_index\fR
make web service with memory\-resident
hashtables  (default=off)
.TP
\fB\-c\fR, \fB\-\-client\fR=\fIhostname\fR:port
run as a client using named host service.
.IP
Copyright (c) 2007\-2008 Michael Casey, Christophe Rhodes
.IP
Goldsmiths, University of London
.SH "SEE ALSO"
The full documentation for
.B audioDB
is maintained as a Texinfo manual.  If the
.B info
and
.B audioDB
programs are properly installed at your site, the command
.IP
.B info audioDB
.PP
should give you access to the complete manual.