2017 Poster Sessions : Augmenting Cloth Simulation with High Frequency Details Learned from Real-World Data

Student Name : Jenny Jin
Advisor : Ron Fedkiw
Research Areas: Graphics/HCI
While researchers have made significant advancement in the area of cloth simulation over the past few decades, the state-of-the-art virtual cloth is still lacking in realism in comparison with the real cloth, which often displays a lot more folds and wrinkles. This problem is especially manifest in coarser cloth meshes, which are most relevant for real-time applications such as gaming and virtual/augmented reality. Thus we propose a new approach to learn the high frequency details from captured real-world data using computer vision and machine learning techniques, and subsequently apply the learned model to a coarser cloth mesh simulated with our new inequality cloth framework which can better incorporate folds and wrinkles, leading to improved realism and visual appeal.

Jenny Jin is a third year PhD candidate in Professor Ron Fedkiw's group at Stanford University in the Computer Science Department. Her main interests are in computer graphics and vision. She is particularly interested in applying computer vision and machine learning techniques improve graphics simulation. She is currently supported by the Stanford Graduate Fellowship. Prior to Stanford, Jenny studied Physics at Princeton University.