Picture for Xiaodan Liang

Xiaodan Liang

Revisiting Over-smoothing in BERT from the Perspective of Graph

Feb 17, 2022
Han Shi, Jiahui Gao, Hang Xu, Xiaodan Liang, Zhenguo Li, Lingpeng Kong, Stephen M. S. Lee, James T. Kwok

Figure 1 for Revisiting Over-smoothing in BERT from the Perspective of Graph

Figure 2 for Revisiting Over-smoothing in BERT from the Perspective of Graph

Figure 3 for Revisiting Over-smoothing in BERT from the Perspective of Graph

Figure 4 for Revisiting Over-smoothing in BERT from the Perspective of Graph

* Accepted by ICLR 2022 (Spotlight)

Access Paper or Ask Questions

Exploring Inter-Channel Correlation for Diversity-preserved KnowledgeDistillation

Feb 08, 2022
Li Liu, Qingle Huang, Sihao Lin, Hongwei Xie, Bing Wang, Xiaojun Chang, Xiaodan Liang

Figure 1 for Exploring Inter-Channel Correlation for Diversity-preserved KnowledgeDistillation

Figure 2 for Exploring Inter-Channel Correlation for Diversity-preserved KnowledgeDistillation

* Accepted by ICCV 2021

Access Paper or Ask Questions

Contrastive Instruction-Trajectory Learning for Vision-Language Navigation

Dec 09, 2021
Xiwen Liang, Fengda Zhu, Yi Zhu, Bingqian Lin, Bing Wang, Xiaodan Liang

Figure 1 for Contrastive Instruction-Trajectory Learning for Vision-Language Navigation

Figure 2 for Contrastive Instruction-Trajectory Learning for Vision-Language Navigation

Figure 3 for Contrastive Instruction-Trajectory Learning for Vision-Language Navigation

Figure 4 for Contrastive Instruction-Trajectory Learning for Vision-Language Navigation

* Accepted by AAAI 2022

Access Paper or Ask Questions

Towards Scalable Unpaired Virtual Try-On via Patch-Routed Spatially-Adaptive GAN

Nov 20, 2021
Zhenyu Xie, Zaiyu Huang, Fuwei Zhao, Haoye Dong, Michael Kampffmeyer, Xiaodan Liang

Figure 1 for Towards Scalable Unpaired Virtual Try-On via Patch-Routed Spatially-Adaptive GAN

Figure 2 for Towards Scalable Unpaired Virtual Try-On via Patch-Routed Spatially-Adaptive GAN

Figure 3 for Towards Scalable Unpaired Virtual Try-On via Patch-Routed Spatially-Adaptive GAN

Figure 4 for Towards Scalable Unpaired Virtual Try-On via Patch-Routed Spatially-Adaptive GAN

* 12 pages, 8 figures, 35th Conference on Neural Information Processing Systems

Access Paper or Ask Questions

IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoning

Nov 07, 2021
Pan Lu, Liang Qiu, Jiaqi Chen, Tony Xia, Yizhou Zhao, Wei Zhang, Zhou Yu, Xiaodan Liang, Song-Chun Zhu

Figure 1 for IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoning

Figure 2 for IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoning

Figure 3 for IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoning

Figure 4 for IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoning

* Accepted to NeurIPS 2021 Track on Datasets and Benchmarks, 27 pages, 18 figures, project available at https://iconqa.github.io

Access Paper or Ask Questions