Generating Soundwaves via Granular Synthesis and Reinforcement Learning
نویسنده
چکیده
We describe an audio granular synthesis generator with controllers that can be accessed by reinforcement learning agents. The movement of the controllers affects the sound, which is analyzed to produce a vallue called the reinforcement. The analysis is based on spectral goals and the reinforcement value is used to adjust the agents. Experiments are described using spectral features that are the spread and centroid, as well as the Mel-Frequency Cepstral Coefficients of the sound. We extend this work to include the complex task of generating a soundwave to match an instrumental recording. We have generated soundwaves that match criteria for reinforcement and gained insight in using MFCCs.
منابع مشابه
Emitting Grains of Learning with Pd
We present work in granular synthesis that combines spectral analysis with reinforcement learning, a type of machine learning. We describe an audio granular synthesis generator designed specifically to perform granular synthesis, with programmable controllers that can be accessed by reinforcement learning controllers. The movement of the controllers affects the sound, which is analyzed to produ...
متن کاملLeveraging Grammar and Reinforcement Learning for Neural Program Synthesis
Program synthesis is the task of automatically generating a program consistent with a specification. Recent years have seen proposal of a number of neural approaches for program synthesis, many of which adopt a sequence generation paradigm similar to neural machine translation, in which sequence-to-sequence models are trained to maximize the likelihood of known reference programs. While achievi...
متن کاملDevelopment of Reinforcement Learning Algorithm to Study the Capacity Withholding in Electricity Energy Markets
This paper addresses the possibility of capacity withholding by energy producers, who seek to increase the market price and their own profits. The energy market is simulated as an iterative game, where each state game corresponds to an hourly energy auction with uniform pricing mechanism. The producers are modeled as agents that interact with their environment through reinforcement learning (RL...
متن کاملRelational Reinforcement Learning via Sampling the Space of First-Order Conjunctive Features
We propose a novel method for reinforcement learning in domains that are best described using relational (“first-order”) features. Our approach is to rapidly sample a large space of such features, selecting a good subset to use as the basis for our Q-function. Our Q-function is created via a regression model that combines the collection of first-order features into a single prediction. To contr...
متن کاملImplementing Real-Time Granular Synthesis
This article describes a flexible architecture for Real-Time Granular Synthesis that accommodates a number of Granular Synthesis variants including Tapped Delay Line, Stored Sample and Synthetic Grain Granular Synthesis in both Pitch Synchronous and Asynchronous forms. Three efficient algorithms for generating grain envelopes are reviewed and two methods for generating stochastic grain onset ti...
متن کامل