A Generative Model for Multi-Dialect Representation

TL;DR

This paper introduces the Mode Synthesizing Machine (MSM), a new generative model designed to effectively represent and learn from handwritten multi-dialect data, outperforming traditional RBMs.

Contribution

The paper proposes the MSM, a novel hierarchical generative model that captures multi-dialect features more efficiently than existing RBMs.

Findings

MSM achieves lower error rates than RBM on handwritten multi-dialect data.

MSM effectively learns and represents multiple dialects in a generative manner.

Experimental results demonstrate MSM's superior performance over RBM.

Abstract

In the era of deep learning several unsupervised models have been developed to capture the key features in unlabeled handwritten data. Popular among them is the Restricted Boltzmann Machines RBM. However, due to the novelty in handwritten multidialect data, the RBM may fail to generate an efficient representation. In this paper we propose a generative model, the Mode Synthesizing Machine MSM for on-line representation of real life handwritten multidialect language data. The MSM takes advantage of the hierarchical representation of the modes of a data distribution using a two-point error update to learn a sequence of representative multidialects in a generative way. Experiments were performed to evaluate the performance of the MSM over the RBM with the former attaining much lower error values than the latter on both independent and mixed data set.