Publications and Papers in Prep

A Study on Improving Realism of Synthetic Data for Machine Learning

Keywords: Synthetic data, Domain gap, Generative adversarial learning, Synthetic-to-real translation

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.

Unsupervised Synthetic Image Refinement via Contrastive Learning and Consistent Semantic-Structural Constraints

Keywords: Synthetic images refinement, Contrastive learning, Image translation, Unsupervised learning, Synthetic to real translation

Ensuring the realism of computer-generated synthetic images is crucial to deep neural network (DNN) training. Due to different semantic distributions between synthetic and real-world captured datasets, there exists semantic mismatch between synthetic and refined images, which in turn results in the semantic distortion. Recently, contrastive learning (CL) has been successfully used to pull correlated patches together and push uncorrelated ones apart. In this work, we exploit semantic and structural consistency between synthetic and refined images and adopt CL to reduce the semantic distortion. Besides, we incorporate hard negative mining to improve the performance furthermore. We compare the performance of our method with several other benchmarking methods using qualitative and quantitative measures and show that our method offers the state-of-the-art performance.

Active Sampling Strategies for Power System Surrogates (In Prep)

Keywords: Data-driven modeling techniques, active learning, machine learning, power system dynamics

With the growing penetration of converter interfaced generation in power systems, the dynamical behavior of these systems is rapidly evolving. One of the challenges with converter-interfaced generation is the increased number of equations, as well as the required numerical timestep, involved in simulating these systems. Within this work, we explore the use of surrogate modeling to cheaply, and accurately, predict the dynamic response of power systems subject to a disturbance for varying system parameters. We consider different active learning strategies that try to minimize the training data requirements for building a sufficiently accurate surrogate.