Cunjun Yu

I am a research assistant at AdaComp Lab advised by Prof David Hsu.

At SenseTime, I was fortunate to have guidance from Prof Hongsheng Li. Previously, I obtained my bachelor's degree at Nanyang Technological University (NTU) and worked closely with Prof Pham Cuong Quong at CRI Group.

Email  /  CV  /  Github  /  Google Scholar  / 

profile photo
invigorate_img INVIGORATE: Interactive Visual Grounding and Grasping in Clutter
Hanbo Zhang*, Yunfan Lu*, Cunjun Yu, David Hsu, Xuguang Lan, Nanning Zheng
RSS, 2021
PDF / Video

This paper presents INVIGORATE, a robot system that interacts with human through natural language and grasps a specified object in clutter.

balms REFINE: Prediction Fusion Network for Panoptic Segmentation
Jiawei Ren*, Cunjun Yu*, Zhongang Cai*, Chongsong Chen, Mingyuan Zhang, Haiyu Zhao, Shuai Yi, Hongsheng Li
AAAI, 2021
PDF / Code / Project Page

We proposed a prediction fusion framework for panoptic segmentation.

balms Balanced Meta-Softmax for Long-Tailed Visual Recognition
Jiawei Ren, Cunjun Yu, Shunan Sheng, Xiao Ma, Haiyu Zhao, Shuai Yi, Hongsheng Li
NeurIPS, 2020
Arxiv / Code / Project Page

We presented Balanced Softmax to accommodate the label distribution shift between training and testing, and a complementary Meta Sampler to estimate the optimal class sample rate.

star Spatio-Temporal Graph Transformer Networks for Pedestrian Trajectory Prediction
Cunjun Yu*, Xiao Ma*, Jiawei Ren, Haiyu Zhao, Shuai Yi,
ECCV, 2020
Arxiv / Code / Project Page / Video

Presented STAR, a Spatio-Temporal grAph tRansformer framework, which captures complex spatio-temporal interactions by interleaving between spatial and temporal Transformers.

star MessyTable: Instance Association in Multiple Camera Views
Zhongang Cai, Junzhe Zhang, Daxuan Ren, Cunjun Yu, Haiyu Zhao, Shuai Yi, Chai Kiat Yeo, Chen Change Loy
ECCV, 2020
Arxiv / Code / Project Page

We presented an interesting and challenging dataset that features a large number of scenes with messy tables captured from multiple camera views.

vsnet Siamese Convolutional Neural Network for Sub-millimeter-accurate Camera Pose Estimation and Visual Servoing
Cunjun Yu*, Zhongang Cai*, Hung Pham, Quang-Cuong Pham
IROS, 2019
Arxiv / Dataset

We proposed a new neural network, based on a Siamese architecture, for highly accurate camera pose estimation.

star Balanced Activation for Long-tailed Visual Recognition
Jiawei Ren*, Cunjun Yu*, Zhongang Cai*, Haiyu Zhao
ECCVW, 2020
Technical Report / Code / Talk / Video

Spotlight of ECCV 2020 COCO + LVIS Workshop.

3d-domainadapt Leveraging Temporal Information for 3D Detection and Domain Adaptation
Cunjun Yu*, Zhongang Cai*, Daxuan Ren, Haiyu Zhao
CVPRW, 2020

3rd Place solution for Waymo Open Dataset Challenge (Domain Adaptation) at CVPR 2020

panopticseg Panoptic HTC with Class-Guided Fusion
Chongsong Chen*, Jiawei Ren*, Daisheng Jin, Zhongang Cai, Cunjun Yu, Bairun Wang, Mingyuan Zhang, Jinyi Wu
ICCVW, 2019
Technical Report / Leader Board (Team Innovation)

1st place solution for ICCV 2019 COCO Panoptic Segmentation Challenge.


Reviewer: IROS 2020, ICRA 2020, NeurIPS 2021, ICLR 2022

Stolen from Jon Barron