Resolving label uncertainty with implicit posterior models
This addresses label uncertainty in machine learning for applications like aerial imagery and text classification, but it appears incremental as it builds on existing concepts of generative models and weak supervision.
The paper tackles the problem of jointly inferring labels across data samples with weak prior beliefs, such as noisy or incomplete labels, by proposing a method that unifies various machine learning settings and demonstrates it on diverse tasks like classification with negative examples and weakly supervised segmentation.
We propose a method for jointly inferring labels across a collection of data samples, where each sample consists of an observation and a prior belief about the label. By implicitly assuming the existence of a generative model for which a differentiable predictor is the posterior, we derive a training objective that allows learning under weak beliefs. This formulation unifies various machine learning settings; the weak beliefs can come in the form of noisy or incomplete labels, likelihoods given by a different prediction mechanism on auxiliary input, or common-sense priors reflecting knowledge about the structure of the problem at hand. We demonstrate the proposed algorithms on diverse problems: classification with negative training examples, learning from rankings, weakly and self-supervised aerial imagery segmentation, co-segmentation of video frames, and coarsely supervised text classification.