Kehan Wang

Kehan Wang

Software engineer working on conversion modeling at Google Ads.

Publications

 Multimodal Semantic Mismatch Detection in Social Media Posts
Kehan Wang, Seth Z. Zhao, David Chan, Avideh Zakhor, John Canny
MMSP, 2022
paper / code

Detect misinformation in Twitter video posts based on multimodal semantic mismatch.

 Neural Face Identification in a 2D Wireframe Projection of a Manifold Object
Kehan Wang, Jia Zheng, Zihan Zhou
CVPR, 2022
project page / arXiv / code

The first data-driven approach to identify face loops from a single line drawing and reconstruct 3D model using face information.

 Multi-Modal Semantic Inconsistency Detection in Social Media News Posts
Scott McCrae, Kehan Wang, Avideh Zakhor
MMM, 2022
paper

Detect misinformation in Facebook news posts in social media using multi-modal fusion.

 Wavelet: Efficient DNN Training with Tick-Tock Scheduling
Guanhua Wang, Kehan Wang, Kenan Jiang, Xiangjun Li, Ion Stoica
MLSys, 2021
project page / paper

Wavelet is a novel scheduling scheme for DNN training that trains model up to 6.7x faster.

 Fast, Accurate Barcode Detection in Ultra High-Resolution Images
Jerome Quenum, Kehan Wang, Avideh Zakhor
ICIP, 2021
paper / code

Y-Net outperforms state-of-the-art methods on barcode detection in high resolution images.

 Composing MPC with LQR and Neural Networks for Efficient and Stable Control
Fangyu Wu, Guanhua Wang, Siyuan Zhuang, Kehan Wang, Alexander Keimer, Ion Stoica, Alexandre Bayen
arXiv, 2021
paper

Memory-Augmented MPC is a provably stable control that combines a linear quadratic regulator, a neural network, and an MPC.