The present disclosure describes techniques for controllable music generation. The techniques comprise extracting latent vectors from unlabelled data, the unlabelled data comprising a plurality of music note sequences, the plurality of music note sequences indicating a plurality of pieces of music; clustering the latent vectors into a plurality of classes corresponding to a plurality of music styles; generating a plurality of labelled latent vectors corresponding to the plurality of music styles, each of the plurality labelled latent vectors comprising information indicating features of a corresponding music style; and generating a first music note sequence indicating a first piece of music in a particular music style among the plurality of music styles based at least in part on a particular labelled latent vector among the plurality of labelled latent vectors, the particular labelled latent vector corresponding to the particular music style.

Apparatus and methods for audio classifying and processing are disclosed. In one embodiment, an audio processing apparatus includes an audio classifier for classifying an audio signal into at least one audio type in real time; an audio improving device for improving experience of audience; and an adjusting unit for adjusting at least one parameter of the audio improving device in a continuous manner based on the confidence value of the at least one audio type.

A musical piece structure analysis method includes acquiring an acoustic signal of a musical piece, extracting a first feature amount indicating changes in tone from the acoustic signal of the musical piece, extracting a second feature amount indicating changes in chords from the acoustic signal of the musical piece, outputting a first boundary likelihood indicating likelihood of a constituent boundary of the musical piece from the first feature amount using a first learning model, outputting a second boundary likelihood indicating likelihood of the constituent boundary of the musical piece from the second feature amount using a second learning model, identifying the constituent boundary of the musical piece by performing weighted synthesis of the first boundary likelihood and the second boundary likelihood, and dividing the acoustic signal of the musical piece into a plurality of sections at the constituent boundary that has been identified.

Embodiments of the present disclosure relate to a sensory experience providing apparatus for providing a sensory experience based on sound in a vehicle, and a method thereof. The controller is configured to receive a sound played in the vehicle, extract a sound feature from the received sound, generate sensory information based on the extracted sound feature, and provide a sensory experience based on the sensory information.

Vehicle machine learning methods include providing one or more computer processors communicatively coupled with a vehicle. Using data gathered from biometric sensors and/or vehicle sensors, a machine learning model is trained to determine a mental state of a driver and/or a driving state corresponding with a portion of a trip. In implementations the mental or driving state may be determined without a machine learning model. Based at least in part on the determined mental state and the determined driving state, one or more interventions are automatically initiated to alter the mental state of the driver. The interventions may include preparing (or modifying) and initiating a music playlist, altering a lighting condition within the vehicle, altering an audio condition within the vehicle, altering a temperature condition within the vehicle, and initiating, altering, or withholding conversation from a conversational agent. Vehicle machine learning systems perform the vehicle machine learning methods.

A method of making audio music selection and creating a mixtape, comprising importing song files from a song repository; sorting and filtering the song files based on selection criteria; and creating the mixtape from the song files sorting and filtering results. The sorting and filtering of the song files comprise: spectral analyzing each of the song files to extract low level acoustic feature parameters of the song file; from the low level acoustic feature parameter values, determining the high level acoustic feature parameters of the analyzed song file; determining a similarity score of each of the analyzed song files by comparing the acoustic feature parameter values of the analyzed song file against desired acoustic feature parameter values determined from the selection criteria; and sorting the analyzed song files according to their similarity scores; and filtering out the analyzed song files with first similarity scores lower than a filter threshold.

An approach is provided in which an information handling system configures a reinforcement learning model based inspiration selections received from a user. The information handling system performs training iterations using the configured reinforcement learning model, which generates multiple actions and multiple rewards corresponding to multiple actions. The information handling system determines that the multiple rewards reach an empirical threshold and, in turn, generates a musical composition based on the multiple actions.

The intelligent accompaniment generating system includes an input module, an analysis module, a generation module and a musical equipment. The input module is configured to receive a musical pattern signal derived from a raw signal. The analysis module is configured to analyze the musical pattern signal to extract a set of audio features, wherein the input module is configured to transmit the musical pattern signal to the analysis module. The generation module is configured to obtain a playing assistance information having an accompaniment pattern from the analysis module, wherein the accompaniment pattern has at least two parts having different onsets therebetween, and each onsets of the at least two parts is generated by an algorithm according to the set of audio features. The musical equipment includes a digital amplifier configured to output an accompaniment signal according to the accompaniment pattern.

Example methods, apparatus and articles of manufacture to determine semantic information for audio are disclosed. Example apparatus disclosed herein are to process an audio signal obtained by a media device to determine values of a plurality of features that are characteristic of the audio signal, compare the values of the plurality of features to a first template having corresponding first ranges of the plurality of features to determine a first score, the first template associated with first semantic information, compare the values of the plurality of features to a second template having corresponding second ranges of the plurality of features to determine a second score, the second template associated with second semantic information, and associate the audio signal with at least one of the first semantic information or the second semantic information based on the first score and the second score.

Digital signal processing and machine learning techniques can be employed in a vocal capture and performance social network to computationally generate vocal pitch tracks from a collection of vocal performances captured against a common temporal baseline such as a backing track or an original performance by a popularizing artist. In this way, crowd-sourced pitch tracks may be generated and distributed for use in subsequent karaoke-style vocal audio captures or other applications. Large numbers of performances of a song can be used to generate a pitch track. Computationally determined pitch trackings from individual audio signal encodings of the crowd-sourced vocal performance set are aggregated and processed as an observation sequence of a trained Hidden Markov Model (HMM) or other statistical model to produce an output pitch track.