Songhua Liu (刘松桦)

A second-year Ph.D. student from National University of Singapore

About Me

Education Background:
- 2021.08 - Now Ph.D.
  Electrical and Computer Engineering
  National Univeristy of Singapore
- 2017.09 - 2021.06 B.S.
  Department of Computer Science and Technology
  Nanjing University
Research Experience:
- 2021.08 - Now
  Learning and Vision Lab (LV Lab)
  National Univeristy of Singapore
  Supervisor: Prof. Xinchao Wang
- 2020.10 - 2021.04
  Department of Computer Vision Technology (VIS)
  Baidu Inc.
  Supervisor: Tianwei Lin
- 2020.02 - 2020.10
  Department of Computer Science
  Rutgers University
  Supervisor: Prof. Hao Wang
- 2019.09 - 2021.06
  State Key Laboratory of New Technology of Computer Software
  Nanjing University
  Supervisor: Prof. Zhengxing Sun
Research Areas:
- Data synthesis in computer vision, such as:
  - Training with Synthetic Samples
  - Dataset Distillation
  - Style Transfer
  - Image Synthesis
  - Artistic Generation
- If you are interested in the above topics, any form of communication and cooperation is welcome. Just drop me an email!
- I believe that AI is a well promising art creator!
Contact: songhua.liu@u.nus.edu

Good News Bad News Normal News

[2022.09.27] One paper rejected by AAAI 2023.
[2022.09.14] Two papers accepted by NeurIPS 2022: Dataset Distillation via Factorization and Deep Model Reassembly (co-author)!
[2022.07.03] Two papers accepted by ECCV 2022: DynaST: Dynamic Sparse Transformer for Exemplar-Guided Image Generation and Learning with Recoverable Forgetting (co-author)!
[2022.06.29] One paper rejected by ACM MM 2022.
[2022.03.01] One paper rejected by CVPR 2022.
[2021.10.18] Invited talk at ICCV 2021 Workshop on AI for Creative Video Editing and Understanding (CVEU). [Bilibili]
[2021.10.14] One paper rejected by AAAI 2022.
[2021.07.23] Two papers accepted by ICCV 2021: Paint Transformer: Feed Forward Neural Painting with Stroke Prediction (Oral) and AdaAttN: Revisit Attention Mechanism in Arbitrary Neural Style Transfer!
[2021.07.03] One paper rejected by ACM MM 2021.
[2021.06.19] Our team, VUE Baidu VIS, got the 1st place on NTIRE 2021 depth guided image relighting challenge, Track II!
[2021.04.21] I feel extra-ordinarily honored to become a Ph.D. student of Prof. Xinchao Wang at NUS! Good good study, day day up in my future Ph. D. career!
[2021.04.16] This site is set up!

Full List

Dataset Distillation via Factorization
Songhua Liu, Kai Wang, Xingyi Yang, Jingwen Ye, Xinchao Wang
NeurIPS 2022 Spotlight
[Paper]|[PyTorch Code]|[Poster]

DynaST: Dynamic Sparse Transformer for Exemplar-Guided Image Generation
Songhua Liu, Jingwen Ye, Sucheng Ren, Xinchao Wang
ECCV 2022
[Paper]|[PyTorch Code]|[Poster]|[Presentation & Slide]

Paint Transformer: Feed Forward Neural Painting with Stroke Prediction
Songhua Liu, Tianwei Lin, Dongliang He, Fu Li, Ruifeng Deng, Xin Li, Errui Ding, Hao Wang
ICCV 2021 Oral
[Paper]|[Paddle Code]|[PyTorch Code]|[Colab Demo]|[Presentation & Slide]
You can also have a try on 一刻相册 or 剪映 app.
AdaAttN: Revisit Attention Mechanism in Arbitrary Neural Style Transfer
Songhua Liu, Tianwei Lin, Dongliang He, Fu Li, Meiling Wang, Xin Li, Zhengxing Sun, Qian Li, Errui Ding
ICCV 2021
[Paper]|[PyTorch Code]|[Online Demo]|[Presentation & Slide]
You can also have a try on 一刻相册 app.
Stable Video Style Transfer Based on Partial Convolution and Depth-Aware Supervision
Songhua Liu, Hao Wu, Shoutong Luo, Zhengxing Sun
ACM MM 2020
[Paper]|[PyTorch Code]|[Presentation]|[Slide]

NUS Research Scholarship, 2021 intake
1st place, NTIRE 2021 depth guided image relighting challenge, Track II
Outstanding winner (特等奖), Undergraduate excellent graduation thesis of Jiangsu Province, 1 out of the whole department, 7 out of the whole university
Finalist, MCM (The Mathematical Contest in Modeling) 2021
Excellent National SRTP Project, 2020
Sun Zhongxiu Scholarship, 2020
NOIP, 2nd Prize, 2016

说话让人温暖，做事问心无愧，问心绝对客观。