Information processing device
An information processing device obtains audio data, obtains, as teacher vibration data, information regarding vibration data used to vibrate a vibrating device, the vibration data being produced on a basis of the audio data, performs machine learning using the audio data and the teacher vibration data as input, and generates learned model data used to convert an audio waveform into a vibration waveform.
1. An information processing device comprising:
an audio data obtaining section configured to obtain audio data;
a teacher vibration data obtaining section configured to obtain, as teacher vibration data, information regarding vibration data used to vibrate a vibrating device, the vibration data being produced on a basis of the audio data; and
a machine learning section configured to perform machine learning using the audio data and the teacher vibration data as input, and generate learned model data used to convert an audio waveform into a vibration waveform.
2. The information processing device according to claim 1 , wherein the machine learning section performs the machine learning using, as an input feature quantity, a component value in each of a plurality of frequency bands obtained by performing frequency analysis of the audio data.
3. The information processing device according to claim 2 , wherein the machine learning section performs the machine learning using an input feature quantity obtained by adding, to the component value in each of the plurality of frequency bands, information regarding a position of the frequency band with respect to a whole.
4. The information processing device according to claim 2 , wherein,
for the component value in each of the plurality of frequency bands as the input feature quantity, the machine learning section outputs two kinds of feature quantities representing a multiplying factor and an addition value for the component value by the machine learning, and
conversion from the audio waveform to the vibration waveform is performed by using the two kinds of feature quantities.
5. The information processing device according to claim 1 , wherein
the audio data obtaining section obtains tag information indicating kinds of vibration together with the audio data, and
the machine learning section refers to the tag information and performs machine learnings independent of each other for the respective kinds of vibration, and generates a plurality of pieces of learned model data respectively associated with the kinds of vibration.
6. The information processing device according to claim 1 , wherein
the teacher vibration data obtaining section obtains information regarding teacher vibration data produced by applying a plurality of kinds of manipulation processing to the audio data, and
the machine learning section performs a plurality of kinds of machine learning using, as teacher data, information regarding vibration data as a result of applying at least a part of the plurality of kinds of manipulation processing, respectively, and generates a plurality of pieces of learned model data respectively associated with the kinds of manipulation processing.
7. The information processing device according to claim 1 , wherein the machine learning section generates a plurality of noise-added conversion models by adding a plurality of kinds of noises different from each other to a conversion model used to convert the audio waveform into the vibration waveform, and performs the machine learning selectively using a part of output feature quantity data close to the teacher vibration data among a plurality of pieces of output feature quantity data obtained by inputting the audio data to each of the plurality of noise-added conversion models.
8. An information processing method comprising:
obtaining audio data;
obtaining, as teacher vibration data, information regarding vibration data used to vibrate a vibrating device, the vibration data being produced on a basis of the audio data; and
performing machine learning using the audio data and the teacher vibration data as input, and generating learned model data used to convert an audio waveform into a vibration waveform.
9. A non-transitory, computer readable storage medium containing a computer program, which when executed by a computer, causes the computer to carry out actions, comprising:
obtaining audio data;
obtaining, as teacher vibration data, information regarding vibration data used to vibrate a vibrating device, the vibration data being produced on a basis of the audio data; and
performing machine learning using the audio data and the teacher vibration data as input, and generating learned model data used to convert an audio waveform into a vibration waveform.
10. A non-transitory, computer readable storage medium containing learned model data used to convert an audio waveform into a vibration waveform, the learned model data being obtained as a result of performing machine learning using audio data and teacher vibration data as input, the teacher vibration data being information regarding vibration data used to vibrate a vibrating device, the vibration data being produced on a basis of the audio data.
11. An information processing device comprising:
a target audio data obtaining section configured to obtain target audio data to be processed; and
a vibration data generating section configured to generate vibration data for vibrating a vibrating device by converting the target audio data into a vibration waveform by using learned model data obtained as a result of performing machine learning using audio data and teacher vibration data as input, the teacher vibration data being information regarding vibration data used to vibrate the vibrating device, the vibration data being produced on a basis of the audio data.