Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Hengshuang Zhao

Hengshuang Zhao

$BT^2$: Backward-compatible Training with Basis Transformation

Nov 08, 2022
Yifei Zhou, Zilu Li, Abhinav Shrivastava, Hengshuang Zhao, Antonio Torralba, Taipeng Tian, Ser-Nam Lim

Figure 1 for $BT^2$: Backward-compatible Training with Basis Transformation

Figure 2 for $BT^2$: Backward-compatible Training with Basis Transformation

Figure 3 for $BT^2$: Backward-compatible Training with Basis Transformation

Figure 4 for $BT^2$: Backward-compatible Training with Basis Transformation

* 13 pages, 2 figures

Access Paper or Ask Questions

Point Transformer V2: Grouped Vector Attention and Partition-based Pooling

Oct 12, 2022
Xiaoyang Wu, Yixing Lao, Li Jiang, Xihui Liu, Hengshuang Zhao

Figure 1 for Point Transformer V2: Grouped Vector Attention and Partition-based Pooling

Figure 2 for Point Transformer V2: Grouped Vector Attention and Partition-based Pooling

Figure 3 for Point Transformer V2: Grouped Vector Attention and Partition-based Pooling

Figure 4 for Point Transformer V2: Grouped Vector Attention and Partition-based Pooling

* Accepted at NeurIPS 2022

Access Paper or Ask Questions

SegPGD: An Effective and Efficient Adversarial Attack for Evaluating and Boosting Segmentation Robustness

Jul 25, 2022
Jindong Gu, Hengshuang Zhao, Volker Tresp, Philip Torr

Figure 1 for SegPGD: An Effective and Efficient Adversarial Attack for Evaluating and Boosting Segmentation Robustness

Figure 2 for SegPGD: An Effective and Efficient Adversarial Attack for Evaluating and Boosting Segmentation Robustness

Figure 3 for SegPGD: An Effective and Efficient Adversarial Attack for Evaluating and Boosting Segmentation Robustness

Figure 4 for SegPGD: An Effective and Efficient Adversarial Attack for Evaluating and Boosting Segmentation Robustness

* European Conference on Computer Vision (ECCV) , 2022

Access Paper or Ask Questions

DecoupleNet: Decoupled Network for Domain Adaptive Semantic Segmentation

Jul 20, 2022
Xin Lai, Zhuotao Tian, Xiaogang Xu, Yingcong Chen, Shu Liu, Hengshuang Zhao, Liwei Wang, Jiaya Jia

Figure 1 for DecoupleNet: Decoupled Network for Domain Adaptive Semantic Segmentation

Figure 2 for DecoupleNet: Decoupled Network for Domain Adaptive Semantic Segmentation

Figure 3 for DecoupleNet: Decoupled Network for Domain Adaptive Semantic Segmentation

Figure 4 for DecoupleNet: Decoupled Network for Domain Adaptive Semantic Segmentation

* Accepted to ECCV 2022. Code is available at https://github.com/dvlab-research/DecoupleNet

Access Paper or Ask Questions

Universal Adaptive Data Augmentation

Jul 14, 2022
Xiaogang Xu, Hengshuang Zhao, Philip Torr

Figure 1 for Universal Adaptive Data Augmentation

Figure 2 for Universal Adaptive Data Augmentation

Figure 3 for Universal Adaptive Data Augmentation

Figure 4 for Universal Adaptive Data Augmentation

* under submission

Access Paper or Ask Questions

Stratified Transformer for 3D Point Cloud Segmentation

Mar 28, 2022
Xin Lai, Jianhui Liu, Li Jiang, Liwei Wang, Hengshuang Zhao, Shu Liu, Xiaojuan Qi, Jiaya Jia

Figure 1 for Stratified Transformer for 3D Point Cloud Segmentation

Figure 2 for Stratified Transformer for 3D Point Cloud Segmentation

Figure 3 for Stratified Transformer for 3D Point Cloud Segmentation

Figure 4 for Stratified Transformer for 3D Point Cloud Segmentation

* Accepted to CVPR2022. Code is avaiable at https://github.com/dvlab-research/Stratified-Transformer

Access Paper or Ask Questions