Procedure And Apparatus For Generating Automatic Replay Of Recordings

• US 20080188964A1
• Filed: 11/09/2005
• Published: 08/07/2008
• Est. Priority Date: 11/09/2004
• Status: Active Grant

• Current Assignee: Bang & Olufsen (Bang & Olufsen a/s)
• Original Assignee: Bang & Olufsen (Bang & Olufsen a/s)
• Inventors: Madsen, John, Bech, Soren
• Granted Patent: US 7,899,564 B2
• Time in Patent Office: Days
• US Class Current: 700/94
• CPC Class Codes: G06F 16/634 Query by example, e.g. query by humming
• CPC Class Codes: G06F 16/639 using playlists
• CPC Class Codes: G06F 16/68 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/683 using metadata automatically derived from the content

Abstract

The invention relates to the automatic generation of playlists based on a collection of digital recordings and a desire to listen to a particular artist or genre, or to re-create the mood generated by a particular recording. The procedure comprises automatic analysis of the musical content expressed as combinations of spectral content and rhythm, classification into genres and establishing an internal database using the musical content as classifiers for retrieval. Furthermore a particular recording not yet forming part of the collection of recordings may be subjected to the same kind of analysis based on the same criteria and be used as the expression of a request for replay of other music from the collection of recordings without conscious or verbalised input.

Claims

1. A procedure for controlling the playback of recordings from a collection of recordings, by inputting a playback order structured as a generally fuzzy expression of general characteristics, said collection of recordings being stored in digital form in at least one location or in a distributed storage system, each recording being classified in a multidimensional parameter space common to all recordings stored, preferably without conscious intervention from the user, according to pre-defined or self-organizing parameters, characterized in that the fuzzy expression is input in the form of one or several seed recordings, said seed recordings being classified in the same multidimensional parameter space used for the collection, said multidimensional classification being compared to combinations of classifications identifying each recording in said collection, the output for playback being a list of access codes for those recordings having parameters fulfilling a multidimensional distance criterion with respect to the seed recordings, the list being used to access sequentially the recordings concerned from the storage system.

2. A procedure according to claim 1, characterized in that the pre-defined parameters are based on running spectral analyses for each recording combined with rhythm analyses.

3. A procedure according to claim 1 or 2, characterized in that the classification of each recording consists in defining the multidimensional vectorial distances between self-organizing clusters obtained by applying running spectral analyses and rhythm analyses for the collection of recordings.

4. A procedure according to claim 3 for adding a recording to the collection, characterized in that the recording is subjected to running spectral analyses for each recording combined with rhythm analyses of the types that were applied to the prior members of the collection, and associating the classification with the access code for the added recording.

5. An apparatus, such as a media player, for performing the procedure according to claim 1, characterized in that it comprises means for inputting digital representations of at least one track of music, means for performing a running spectrum analysis on said representation, means for performing a rhythm analysis on said representation, means for entering the results of said analyses in a storage matrix, means for clustering the information in order to obtain a vector representation in a further storage matrix, means for associating said vector representation with meta-data representing discographical information relating to the track of music (artist, composer, title, duration), means for storing and organizing this information in database form.

6. An apparatus according to claim 5, characterized in that it comprises means for inputting digital representations of a large number of tracks of music, means for performing a running spectrum analysis on said representation, means for performing a rhythm analysis on said representation, means for entering the results of said analyses in a storage matrix, means for clustering the information in order to obtain a vector representation in a further storage matrix, means for associating said vector representation with meta-data representing discographical information relating to the tracks of music (artists, composers, titles, durations), means for presenting this information in database form.

7. An apparatus performing the procedure according to claim 1, characterized in that it comprises the apparatus defined in claim 5 for inputting further tracks of music into a database obtained by the apparatus defined in claim 6.

8. An apparatus according to claim 6, characterized in that it further comprises means for entering a query to said database, means for extracting response information relative to said query, means for structuring the response information in the form of a playlist of tracks of music, and means for using said playlist information as an order for consecutive replay of said tracks of music via a replay system.

9. An apparatus according to claim 8, characterized in that the query is input as a complete track of music constituting an exemplary recording by means of a query order linked to the input apparatus defined in claim 5 and working on one individual track of music only.

10. An apparatus performing the procedure according to claim 1 in a manner that causes the generation of a revised playlist during the execution of a track that is presented on an existing playlist, characterized in that an active input to the apparatus is interpreted by the apparatus as an order to generate a playlist of tracks that share characteristics with the executed track, using already determined characteristics for the multidimensional distance criterion to be applied for the new playlist.

11. An apparatus according to any of the claims 5-9, characterized in that the digital representation is a linear PCM representation.

12. An apparatus according to any of the claims 5-9, characterized in that the digital representation is a data reduced representation based on psychoacoustic masking.