Group DETR V2

Group DETR v2 represents a significant advancement in object detection, aiming to improve the accuracy and efficiency of transformer-based detectors like DETR. Current research focuses on refining DETR architectures through techniques such as encoder-decoder pretraining, dynamic query selection, and improved loss functions to address issues like slow convergence and suboptimal performance on small or occluded objects. These improvements have led to state-of-the-art results on benchmark datasets like COCO, demonstrating the potential of Group DETR v2 and related approaches for various computer vision applications, including autonomous driving and medical image analysis.

Papers

April 14, 2023

DETR with Additional Global Aggregation for Cross-domain Weakly Supervised Object Detection
Zongheng Tang, Yifan Sun, Si Liu, Yi Yang
Weak Supervision Weakly Supervised Group DETR V2 Cross Domain Object Detection Image Prediction Global Aggregation

March 2, 2023

FeatAug-DETR: Enriching One-to-Many Matching for DETRs with Feature Augmentation
Rongyao Fang, Peng Gao, Aojun Zhou, Yingjie Cai, Si Liu, Jifeng Dai, Hongsheng Li
Object Detection Feature Augmentation Self Augmentation Group DETR V2

November 22, 2022

DETRs with Collaborative Hybrid Assignments Training
Zhuofan Zong, Guanglu Song, Yu Liu
Group DETR V2 DETR Training Collaborative Training Non Maximum Suppression

November 7, 2022

Group DETR v2: Strong Object Detector with Encoder-Decoder Pretraining
Qiang Chen, Jian Wang, Chuchu Han, Shan Zhang, Zexian Li, Xiaokang Chen, Jiahui Chen, Xiaodi Wang, Shuming Han, Gang Zhang, Haocheng Feng, Kun Yao, Junyu Han, Errui Ding, Jingdong Wang
Vision Transformer Object Detection Group DETR V2 DETR Training ViT Encoder Encoder DEcoder

July 30, 2022

Meta-DETR: Image-Level Few-Shot Detection with Inter-Class Correlation Exploitation
Gongjie Zhang, Zhipeng Luo, Kaiwen Cui, Shijian Lu, Eric P. Xing
Data Detection Shot Object Detection Group DETR V2 Shot Detection Attribute Correlation Efficient Meta Learning Algorithm

March 7, 2022

DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection
Hao Zhang, Feng Li, Shilong Liu, Lei Zhang, Hang Su, Jun Zhu, Lionel M. Ni, Heung-Yeung Shum
End to End Multi Scale Feature Multi Object Object Box Group DETR V2 DiNO Mix

March 2, 2022

D^2ETR: Decoder-Only DETR with Computationally Efficient Cross-Scale Attention
Junyu Lin, Xiaofeng Mao, Yuefeng Chen, Lei Xu, Yuan He, Hui Xue
Multi Scale Encoder Decoder Architecture MEG Decoder Group DETR V2 Cross Scale Attention

January 28, 2022

DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETR
Shilong Liu, Feng Li, Hao Zhang, Xiao Yang, Xianbiao Qi, Hang Su, Jun Zhu, Lei Zhang
Detection Transformer Transformer Decoder Query Expansion Group DETR V2 DETR Based Detector Adaptive Sparse Anchor