Automatic sequencing of video playlists based on mood classification of each video and video cluster transitions

摘要

A given set of videos are sequenced in an aesthetically pleasing manner using models learned from human curated playlists. Semantic features associated with each video in the curated playlists are identified and a first order Markov chain model is learned from curated playlists. In one method, a directed graph using the Markov model is induced, wherein sequencing is obtained by finding the shortest path through the directed graph. In another method a sampling based approach is implemented to produce paths on the digraph. Multiple samples are generated and the best scoring sample is returned as the output. In a third method, a relevance based random walk sampling algorithm is modified to produce a reordering of the playlist.

主权项

1. A method for sequencing videos in a playlist, the method comprising:
extracting mood descriptors from adjectives associated with playlists in a video repository, the mood descriptors providing a summary of musical attributes of most videos of each playlist; identifying curated playlists among the playlists in the video repository, the curated playlists being ordered by a user; training a classifier from the extracted mood descriptors; generating dimensional mood features for each video in the curated playlists based on the trained classifier; clustering the dimensional mood features; mapping each video in each curated playlist to a cluster; generating a transition matrix describing a frequency of transitions between the clusters for each curated playlist; and sequencing a given playlist using at least one of digraph routing, sampling and random walk sequencing methods and based on the generated transition matrix.