Song Han

EfficientViT: Enhanced Linear Attention for High-Resolution Low-Computation Visual Recognition

May 29, 2022
Han Cai, Chuang Gan, Song Han

Figure 2 for EfficientViT: Enhanced Linear Attention for High-Resolution Low-Computation Visual Recognition

Figure 3 for EfficientViT: Enhanced Linear Attention for High-Resolution Low-Computation Visual Recognition

Figure 4 for EfficientViT: Enhanced Linear Attention for High-Resolution Low-Computation Visual Recognition

Access Paper or Ask Questions

BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation

May 26, 2022
Zhijian Liu, Haotian Tang, Alexander Amini, Xinyu Yang, Huizi Mao, Daniela Rus, Song Han

Figure 2 for BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation

Figure 3 for BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation

Figure 4 for BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation

* The first two authors contributed equally to this work. Project page: https://bevfusion.mit.edu

Access Paper or Ask Questions

Lite Pose: Efficient Architecture Design for 2D Human Pose Estimation

May 03, 2022
Yihan Wang, Muyang Li, Han Cai, Wei-Ming Chen, Song Han

Figure 2 for Lite Pose: Efficient Architecture Design for 2D Human Pose Estimation

Figure 3 for Lite Pose: Efficient Architecture Design for 2D Human Pose Estimation

Figure 4 for Lite Pose: Efficient Architecture Design for 2D Human Pose Estimation

* IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR) 2022
* 11 pages

Access Paper or Ask Questions

PVNAS: 3D Neural Architecture Search with Point-Voxel Convolution

Apr 26, 2022
Zhijian Liu, Haotian Tang, Shengyu Zhao, Kevin Shao, Song Han

Figure 2 for PVNAS: 3D Neural Architecture Search with Point-Voxel Convolution

Figure 3 for PVNAS: 3D Neural Architecture Search with Point-Voxel Convolution

Figure 4 for PVNAS: 3D Neural Architecture Search with Point-Voxel Convolution

* Journal extension of arXiv:1907.03739 and arXiv:2007.16100 (IEEE TPAMI, 2021). The first two authors contributed equally to this work

Access Paper or Ask Questions

Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications

Apr 25, 2022
Han Cai, Ji Lin, Yujun Lin, Zhijian Liu, Haotian Tang, Hanrui Wang, Ligeng Zhu, Song Han

Figure 2 for Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications

Figure 3 for Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications

Figure 4 for Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications

* ACM Transactions on Design Automation of Electronic Systems (TODAES), Volume 27, Issue 3, Article 20, Page 1-50, 2021
* Journal preprint (ACM TODAES, 2021). The first seven authors contributed equally to this work and are listed in the alphabetical order

Access Paper or Ask Questions

TorchSparse: Efficient Point Cloud Inference Engine

Apr 21, 2022
Haotian Tang, Zhijian Liu, Xiuyu Li, Yujun Lin, Song Han

Figure 2 for TorchSparse: Efficient Point Cloud Inference Engine

Figure 3 for TorchSparse: Efficient Point Cloud Inference Engine

Figure 4 for TorchSparse: Efficient Point Cloud Inference Engine

* MLSys 2022. The first three authors contributed equally to this work. Project page: https://torchsparse.mit.edu

Access Paper or Ask Questions

On-chip QNN: Towards Efficient On-Chip Training of Quantum Neural Networks

Feb 26, 2022
Hanrui Wang, Zirui Li, Jiaqi Gu, Yongshan Ding, David Z. Pan, Song Han

Figure 2 for On-chip QNN: Towards Efficient On-Chip Training of Quantum Neural Networks

Figure 3 for On-chip QNN: Towards Efficient On-Chip Training of Quantum Neural Networks

Figure 4 for On-chip QNN: Towards Efficient On-Chip Training of Quantum Neural Networks

* Published as a conference paper in DAC 2022; 7 pages, 8 figures; open-source at https://github.com/mit-han-lab/torchquantum

Access Paper or Ask Questions