Signal Search Software introduces Vishesh Geet Mala - a stand alone or network
music on demand product to Indian Market.
The stand alone system is a Digital Juke Box having a set top box with Television Interface.
In addition, Signal Search recommends Song Packages
Beeti Hui Yaadein (Songs of 50's)
Bhanware Ki Gunjan (Songs of 60's)
Nasheelee Shaam (Songs of 70's)
Geeton Ke Mehfil (Songs of 80's)
Suron Ke Sargam (Songs of 90's)
Aati Kya Khandala ? (Songs of 2000's)
Signal Search Software introduces Gunjan the mobile application to access
Bollywood songs that can be streamed directly from YouTube to mbile phones
The Matching Algorithms developed by Signal Search Software assoicates her Database
of Metadata with YouTube Content in a machine to machine communication using
the YouTube APIs.
Gunjan For Nokia Mobile Phones
Gunjan For Blackberry
Gunjan For Android
Gunjan For Iphones
The product framework ensures
i) Constant availability of music anytime / anywhere on network
Wired, Wireless, Wide Area, Local Area, Mobile, Satellite, Cable
ii) A multimodal search capability of content
iii) Streaming of selected content on network
iv) Negotiable content
|The basic edition of Vishesh Geet Mala carries only Bollywood Hindi Songs from 1950 onwards.|
Vishesh Geet Mala has three basic components, namely
i) User Access To Search and Experience Music Content - the Client
ii) Search / Retrieval Engine - on Server
iii) Delivery on Network of Selected Content - Streaming from Server
The User Access (UA) and Delivery on Network (DN) depend on nature of network
such as Internet, Cable Television and Mobile. The features offered by the
Search / Retrieval Engine are also dependent on nature of network.
The engine operates on a large database of metadata, reduced representation of
and original content, offering multimodal facility,by way of
Text Based Search
By title of song,
name of singer,
name of lyricist,
name of music director,
name of movie,
year or range of years in which movie was released,
name of actor.
Other features such as album, producer etc, can be introduced if desired.
Query by Hum
The user may hum a tune and record the same using any sound recorder such as the
open source sound editor / recorder Audacity, which is readily downloadable,as per
known format such as wav file. The hummed tune can be uploaded as a query for the
search engine to process.
The current implementation imposes a certain style of humming - for example
using "ta" syllable for open mouth and "Mm" for closed mouth humming.
A demo of how this may be done will be made available on our website
The algorithms of Signal Search Software are modelled after the way we as humans recall tunes.
But as one may expect, the closer the hum to original tune,the better the hits will be.
The software provides a user interface to set confidence level parameter for measuring level of
accuracy. A good singer may set the parameter to 90% and above whereas a bad singer may
lower the parameter to 10%.
Future UpgradesSong Recognition from Sample / Signature
A recording of a portion of a song or signature tune of a movie can be uploaded
for the search engine to look for the song/movie. This feature will be made
available in the third release. The complete metadata of the song / movie will
be output by the system.
Song Recognition from Spoken Content / Singing
This feature will be based on Signal Search Speech / Tune Recognition Module.
It will be a trainable software that can be downloaded from Signal Search Website.