公开(公告)号：WO2020157283A1

公开(公告)日：2020-08-06

申请号：PCT/EP2020/052446

申请日：2020-01-31

Applicant: MOODAGENT A/S

Inventor： STEFFENSEN, Peter, Berg ,  HENDERSON, Mikael

IPC: G06F16/783 ,  G06F16/683 ,  G06F16/632

Abstract: A method of recommending video content using a computer-based system (30), the method comprising providing (101) an initial set comprising a plurality of videos (1); extracting (102) a digital audio signal (2) from each of the plurality of videos (1); determining (103) at least one temporal sequence (4) of low-level audio features for each digital audio signal (2) of the plurality of videos (1) by analyzing the digital audio signals (2); calculating (104) an audio similarity index (5) between each of the plurality of videos (1) by comparing their respective at least one temporal sequence (4) of low-level audio features; receiving (105) a query Q comprising reference to a seed video; the seed video being one of the plurality of videos (1); determining (106), for the seed video, a ranking (7) of the rest of the initial set of videos (1) based on their audio similarity index (5) with respect to the seed video; and returning (107), as a reply to the query Q , an ordered set of video references according to the ranking (7).

公开(公告)号：WO2020084070A1

公开(公告)日：2020-04-30

申请号：PCT/EP2019/079058

申请日：2019-10-24

Applicant: MOODAGENT A/S

Inventor： DYRSTING, Søren ,  HENDERSON, Mikael ,  STEFFENSEN, Peter, Berg

IPC: G10H1/00

Abstract: A method of determining on a computer-based system at least one representative segment of a musical composition, the method comprising providing (101) a digital audio signal (1) representing said musical composition; dividing (102) said digital audio signal (1) into a plurality of frames (2) of equal frame duration; calculating (103) at least one audio feature value for each frame by analyzing the digital audio signal (1), said audio feature being a numerical representation of a musical characteristic of said digital audio signal (1), with a numerical value equal to or higher than zero; identifying (104) at least one representative frame (3) corresponding to a maximum value of said audio feature; and determining (105) at least one representative segment (4) of the digital audio signal (1) with a predefined segment duration, the starting point of said at least one representative segment (4) being a representative frame (3).

公开(公告)号：WO2021198087A1

公开(公告)日：2021-10-07

申请号：PCT/EP2021/057969

申请日：2021-03-26

Applicant: MOODAGENT A/S

Inventor： STEFFENSEN, Peter, Berg ,  HENDERSON, Mikael ,  JENSEN, Nick

IPC: G10L21/0364 ,  G06F16/683 ,  G10H1/06 ,  H03G5/16 ,  H04S7/00 ,  G10L25/48 ,  G06F3/16 ,  G10H1/12 ,  H04R3/00 ,  G10H2220/371 ,  G10H2240/085 ,  H03G5/005 ,  H03G5/165 ,  H04R2420/03

Abstract: A method and system for optimizing audio playback by dynamically equalizing an audio signal (1), using an associated high-level feature vector (2) comprising high- level feature values (3) representing semantic characteristics of the audio signal (1), for determining a frequency response profile (4) and applying the frequency response profile (4) to the audio signal (1) to produce an equalized audio signal (7) for playback through an audio interface (26).

公开(公告)号：WO2021064026A1

公开(公告)日：2021-04-08

申请号：PCT/EP2020/077377

申请日：2020-09-30

Applicant: MOODAGENT A/S

Inventor： HENDERSON, Mikael ,  STEFFENSEN, Peter, Berg ,  TEGLBJÆRG, David, Stubbe ,  ANDERSEN, Jesper, Steen ,  ANTICH, Jose, Luis, Diez ,  JØRGENSEN, Thomas

IPC: G06F16/68

Abstract: A method, device (1) and system (15) for organizing media content on a computer-based system to form a playlist, wherein the device (1) has access to a database (3) with a plurality of music tracks (4) and associated feature vectors (5) comprising feature values (5A) representing different semantic characteristics of a music track (4), as well as metadata (6) comprising at least one type of metadata record (6A) representing associated information about the respective music track (4). The playlist is determined based on a query (8) from the client device (1) that comprises an input vector (9), and at least one input metadata record (6B), using an additional similarity matrix (7) representing a measure of similarity between different metadata records (6A) of the same type.