Training Better Deep Neural Networks with Uncertainty Mining Net
Yang Sun, Abhishek Kolagunda, Steven Eliuk and Xiaolong Wang*, BMVC, 2021
Training better deep learning model with noise data.
Publications
This page only lists recent publications, for the full list, please go to Google Scholar. (* indicates corresponding author)
Continual Learning Long Short Term Memory
Xin Guo, Yu Tian, Qinghan Xue, Panos Lampropoulos, Steven Eliuk, Kenneth Barner and Xiaolong Wang*, EMNLP(findings), 2020
This is one of the pioneers in continual learning for NLP related works via recurrent neural networks.
AWDF: An Adaptive Weighted Deep Fusion Architecture for Multi-modality Learning
Qinghan Xue, Abhishek Kolagunda, Steven Eliuk, Xiaolong Wang*, IEEE Big Data, 2019
Link paper
Talking Face Generation by Conditional Recurrent Adversarial Network
Yang Sun, Jingwen Zhu, Dawei Li, Xiaolong Wang*, Hairong Li, IJCAI, 2019
About talking face generation framework, we completed the first version in 2017.
Boundary-sensitive Network for Portrait Segmentation
Xianzhi Du, Xiaolong Wang*, Dawei Li, Jingwen Zhu, Serafettin Tasci, Cameron Upright, Stephen Walsh, Larry Davis, FG(oral), 2019
Focusing on the most challenging part of portrait segmentation, we build an end-to-end solution and obtain more realistic results, especially on the object boundary.
Link paper
Are You Speaking: Real-Time Speech Activity Detection via Landmark Pooling Network
Boyu Wang, Xiaolong Wang*, FG, 2019
Real time speech activity detection
MICIK: MIning Cross-Layer Inherent Similarity Knowledge for Deep Model Compression
Jie Zhang, Xiaolong Wang*, Dawei Li, Shalini Ghosh, Abhishek Kolagunda, Yalin Wang, arXiv, 2019
The first cross layers based Deep model compression work.
Link paper
Accelerating Proposal Generation Network for Fast Face Detection on Mobile Devices
Heming Zhang, Xiaolong Wang*, Jingwen Zhu, C.-C. Jay Kuo, ICIP, 2018
Real time unconstrained face detection on mobile.
Link paper
LCANet: End-to-End Lipreading with Cascaded Attention-CTC
Kai Xu, Dawei Li, Nick Cassimatis, Xiaolong Wang*, FG, 2018
One of the earliest works in Lipreading. No only achieving the top performance, but also we make it run real time on mobile devices.
Link paper
Dynamically Hierarchy Revolution: DirNet for Compressing Recurrent Neural Network on Mobile Devices
Jie Zhang, Xiaolong Wang*, Dawei Li, Yalin Wang, IJCAI(oral), 2018
Model compression for recurrent neural network
DeepRebirth: Accelerating Deep Neural Network Execution on Mobile Devices
Dawei Li, Xiaolong Wang*, Deguang Kong, AAAI(oral), 2018
We completed the first pipeline in 2016. THis is one of the pioneers in model compression works.
Link paper