Mengxiao Lin (林孟潇)

I am a researecher at Megvii Technology (Face++). Previously I was working on natural language processing with deep learning approaches, advised by Prof. Xiaoqing Zheng at Fudan University.

I am working at Basemodel Group at Megvii Research. Our work at Face++ is about basemodel training and all major computer vision tasks like object detection, pose estimation, semantic segmentation and so on.

Publications

• A Simple yet Effective Joint Training Method for Cross-Lingual Universal Dependency Parsing
  Danlu Chen^*, Mengxiao Lin^*, Zhifeng Hu^* and Xipeng Qiu.
  In Proceedings of the CoNLL 2018 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies, 2018. (^*: equal contribution).
  
  
• ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices
  Xiangyu Zhang, Xinyu Zhou, Mengxiao Lin and Jian Sun.
  In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018.
  
  
• Context-Specific and Multi-Prototype Character Representations
  Xiaoqing Zheng, Jiangtao Feng, Mengxiao Lin and Wenqiang Zhang.
  In Proceedings of the Twenty-Fifth International Joint Conference on Artificial Intelligence(IJCAI-16), 2016.

Open-source Projects

• Yet Another MXnet DETection
  A Faster R-CNN implementation fully in MXNet gluon API. More functions are in developing.
  
  
• parserChiang
  Naïve transition-based dependency parser in MXNet gluon.