Neural separation of observed and unobserved distributions
This addresses a key challenge in machine learning for scenarios with partial observation, offering a practical solution for signal processing applications.
The paper tackles the problem of separating mixed distributions when only one source distribution is observed, introducing Neural Egg Separation to iteratively learn separation and Latent Mixture Masking for initialization. Experiments on audio and image tasks show it outperforms methods with similar supervision and often matches fully supervised performance.
Separating mixed distributions is a long standing challenge for machine learning and signal processing. Most current methods either rely on making strong assumptions on the source distributions or rely on having training samples of each source in the mixture. In this work, we introduce a new method---Neural Egg Separation---to tackle the scenario of extracting a signal from an unobserved distribution additively mixed with a signal from an observed distribution. Our method iteratively learns to separate the known distribution from progressively finer estimates of the unknown distribution. In some settings, Neural Egg Separation is initialization sensitive, we therefore introduce Latent Mixture Masking which ensures a good initialization. Extensive experiments on audio and image separation tasks show that our method outperforms current methods that use the same level of supervision, and often achieves similar performance to full supervision.