Synbols: Probing Learning Algorithms with Synthetic Datasets

TL;DR

Synbols is a versatile synthetic dataset generator that helps researchers systematically test and analyze the limitations of various machine learning algorithms across multiple learning paradigms.

Contribution

We introduce Synbols, a new tool for creating customizable synthetic datasets to probe and understand learning algorithms' strengths and weaknesses.

Findings

Identified specific failure modes of algorithms using Synbols datasets

Demonstrated the tool's ability to generate diverse and complex data distributions

Showcased insights into algorithm robustness across different learning setups

Abstract

Progress in the field of machine learning has been fueled by the introduction of benchmark datasets pushing the limits of existing algorithms. Enabling the design of datasets to test specific properties and failure modes of learning algorithms is thus a problem of high interest, as it has a direct impact on innovation in the field. In this sense, we introduce Synbols -- Synthetic Symbols -- a tool for rapidly generating new datasets with a rich composition of latent features rendered in low resolution images. Synbols leverages the large amount of symbols available in the Unicode standard and the wide range of artistic font provided by the open font community. Our tool's high-level interface provides a language for rapidly generating new distributions on the latent features, including various types of textures and occlusions. To showcase the versatility of Synbols, we use it to dissect the limitations and flaws in standard learning algorithms in various learning setups including supervised learning, active learning, out of distribution generalization, unsupervised representation learning, and object counting.