Synthetic Data for Model Selection

TL;DR

This paper investigates the use of synthetic data for model selection in image classification, demonstrating its potential to replace validation sets and proposing a calibration method to improve its effectiveness.

Contribution

It introduces a novel approach to use synthetic data for model selection and a calibration technique to align synthetic error estimates with real data.

Findings

Synthetic data can replace validation sets when data is scarce.

Calibration improves the accuracy of synthetic error estimates.

Synthetic data enhances model selection in limited-data scenarios.

Abstract

Recent breakthroughs in synthetic data generation approaches made it possible to produce highly photorealistic images which are hardly distinguishable from real ones. Furthermore, synthetic generation pipelines have the potential to generate an unlimited number of images. The combination of high photorealism and scale turn synthetic data into a promising candidate for improving various machine learning (ML) pipelines. Thus far, a large body of research in this field has focused on using synthetic images for training, by augmenting and enlarging training data. In contrast to using synthetic data for training, in this work we explore whether synthetic data can be beneficial for model selection. Considering the task of image classification, we demonstrate that when data is scarce, synthetic data can be used to replace the held out validation set, thus allowing to train on a larger dataset. We also introduce a novel method to calibrate the synthetic error estimation to fit that of the real domain. We show that such calibration significantly improves the usefulness of synthetic data for model selection.