A Study on Improving Realism of Synthetic Data for Machine Learning
This work addresses the need for more realistic synthetic data to improve machine learning models, but it is incremental as it builds on existing adversarial training methods without introducing a new paradigm.
The study tackled the problem of improving the realism of synthetic data for machine learning by training a synthetic-to-real generative model, resulting in enhanced performance on downstream perception tasks through qualitative and quantitative evaluations.
Synthetic-to-real data translation using generative adversarial learning has achieved significant success in improving synthetic data. Yet, limited studies focus on deep evaluation and comparison of adversarial training on general-purpose synthetic data for machine learning. This work aims to train and evaluate a synthetic-to-real generative model that transforms the synthetic renderings into more realistic styles on general-purpose datasets conditioned with unlabeled real-world data. Extensive performance evaluation and comparison have been conducted through qualitative and quantitative metrics and a defined downstream perception task.