Dynamic R-CNN: Towards High Quality Object Detection via Dynamic Training

The proposed Dynamic R-CNN method for object detection, which adjusts the label assignment criteria and the shape of regression loss function based on the statistics of proposals during training.

Data and Resources

Cite this as

Hongkai Zhang, Hong Chang, Bingpeng Ma, Naiyan Wang, Xilin Chen (2024). Dataset: Dynamic R-CNN: Towards High Quality Object Detection via Dynamic Training. https://doi.org/10.57702/cgy0bfcb

DOI retrieved: December 16, 2024

Additional Info

Field Value
Created December 16, 2024
Last update December 16, 2024
Author Hongkai Zhang
More Authors
Hong Chang
Bingpeng Ma
Naiyan Wang
Xilin Chen
Homepage https://github.com/hkzhang95/DynamicRCNN