Embodiments are disclosed for generating an audio segmentation of an audio sequence using deep embeddings. In particular, in one or more embodiments, the disclosed systems and methods comprise receiving an input including an audio sequence and extracting features for each frame of the audio sequence, where each frame is associated with a beat of the audio sequence. The method may further comprise clustering frames of the audio sequence into one or more clusters based on the extracted features and generating segments of the audio sequence based on the clustered frames, where each segment includes frames of the audio sequence from a same cluster. The method may further comprise constructing a multi-level audio segmentation of the audio sequence and performing a segment fusioning process that merges shorter segments with neighboring segments based on cluster assignments.

Techniques are disclosed relating to generating music content. In one embodiment, a method includes determining one or more musical attributes based on external data and generating music content based on the one or more musical attributes. Generating the music content may include selecting from stored sound loops or tracks and/or generating new tracks based on the musical attributes. Selected or generated sound loops or tracks may be layered to generate the music content. Musical attributes may be determined in some embodiments based on user input (e.g., indicating a desired energy level), environment information, and/or user behavior information. Artists may upload tracks, in some embodiments, and be compensated based on usage of their tracks in generating music content. In some embodiments, a method includes generating sound and/or light control information based on the musical attributes.

A computer-implemented method of generating a musical composition containing a plurality of musical texture groups is disclosed. The method includes assembling musical texture groups from musical instrument components and associating therewith a tag expressing emotional textural connotation. The instrument components have musical textural classifiers selected from a set of pre-defined textural classifiers such that different instrument components may have a different subset of pre-defined textural classifiers. The textural classifiers within a texture group possess either no musical feature attribute or a single musical feature attribute and any number of musical accompaniment attributes. The method then generates at least one chord scheme to a narrative brief, to provide an emotional connotation to a series of events, the chord scheme generated by selecting and assembling Form Atoms. The final step includes applying a texture to the chord scheme to generate the musical composition reflecting the narrative brief.

In an embodiment, a method is provided for determining a mood profile of media data. In this method, mood is determined across multiple elements of mood for the media data to create a mood profile associated with the media data. In some embodiments, the mood profile is then used to determine congruencies between one or more pieces of media data.

The present invention relates to a computer implemented system and method for generating an audio output file. The method including using one or more processors to perform steps of: receiving audio tracks, each audio track created according to audio parameters; separating each audio track into at least one selectable audio block, each audio block including audio content from a musical instrument involved in creating the audio track; assigning a unique identifier to each audio block; using the unique identifiers to select audio blocks, and generating the audio output by combining the audio blocks. The present invention prevents the use of the same combination of audio blocks in the generation of audio output to ensure that the audio output files generated a sufficiently unique. Also provided are audio file recording, editing and mixing modules enabling a user to have full creative control over mix and other parameters to modify as desired the audio file generated.

Example embodiments relate to methods and systems for playback of adaptive music corresponding to an athletic activity. A user input is received from a user selecting an existing song for audible playback to the user, the song comprising a plurality of audio layers including at least a first layer, a second layer, and a third layer. Augmented playback of the existing song to the user is initiated by audibly providing the first layer but not the second layer. Physical activity information derived from a sensor corresponding to a real-time physical activity level of a user is received. If the physical activity level of the user is above a first activity level threshold, the augmented playback of the existing song is continued by audibly providing the first layer and the second layer to the user.

A generative composition system reduces existing musical artefacts to constituent elements termed “Form Atoms”. These Form Atoms may each be of varying length and have musical properties and associations that link together through Markov chains. To provide myriad new composition, a set of heuristics ensures that musical textures between concatenated musical sections follow a supplied and defined briefing narrative for the new composition whilst contiguous concatenated Form Atoms are also automatically selected to see that similarities in respective and identified attributes of musical textures for those musical sections are maintained to maintain good musical form. Within the composition work, chord spacing and control are practiced to maintain musical sense in the new composition and a primitive heuristics structure maintains pitch and permits key transformation. The system provides signal analysis and music generation by allowing emotional connotations to be specified and reproduced from cross-referenced Form-Atoms.

A generative composition system reduces existing musical artefacts to constituent elements termed “Form Atoms”. These Form Atoms may each be of varying length and have musical properties and associations that link together through Markov chains. To provide myriad new composition, a set of heuristics ensures that musical textures between concatenated musical sections follow a supplied and defined briefing narrative for the new composition whilst contiguous concatenated Form Atoms are also automatically selected to see that similarities in respective and identified attributes of musical textures for those musical sections are maintained to support maintenance of musical form. Independent aspects of the disclosure further ensure that, within the composition work, such as a media product or a real-time audio stream, chord spacing determination and control are practiced to maintain musical sense in the new composition. Further, a structuring of primitive heuristics operates to maintain pitch and permit key transformation. The system and its functionality provides signal analysis and music generation through allowing emotional connotations to be specified and reproduced from cross-referenced Form-Atoms.

According to a first embodiment, one method presented herein involves methods of finding the best fitting song from a large audio database for a selected video production. The content of these songs utilized in the instant invention has been tagged by emotion tags describing the energy, the emotion of these songs over time—meaning that each song can contain a plurality of, even overlapping, emotions.

According to a first embodiment, there is presented herein a method of increasing the energy level of a user-selected song in a loop-based music generation system. In one embodiment the algorithm is integrated into a music generation/song construction process and comprises of three different approaches, with one of the three being a hybrid version of the remaining approaches.