CVAIMar 2, 2022

DN-DETR: Accelerate DETR Training by Introducing Query DeNoising

Tsinghua
arXiv:2203.01305v31116 citationsh-index: 80Has Code
Originality Highly original
AI Analysis

This addresses a key bottleneck in training efficiency for DETR-based object detection models, offering a universal plug-in solution.

The paper tackles the slow convergence issue in DETR-like object detection methods by introducing a denoising training technique, resulting in a 1.9 AP improvement and achieving state-of-the-art results with fewer training epochs.

We present in this paper a novel denoising training method to speedup DETR (DEtection TRansformer) training and offer a deepened understanding of the slow convergence issue of DETR-like methods. We show that the slow convergence results from the instability of bipartite graph matching which causes inconsistent optimization goals in early training stages. To address this issue, except for the Hungarian loss, our method additionally feeds ground-truth bounding boxes with noises into Transformer decoder and trains the model to reconstruct the original boxes, which effectively reduces the bipartite graph matching difficulty and leads to a faster convergence. Our method is universal and can be easily plugged into any DETR-like methods by adding dozens of lines of code to achieve a remarkable improvement. As a result, our DN-DETR results in a remarkable improvement ($+1.9$AP) under the same setting and achieves the best result (AP $43.4$ and $48.6$ with $12$ and $50$ epochs of training respectively) among DETR-like methods with ResNet-$50$ backbone. Compared with the baseline under the same setting, DN-DETR achieves comparable performance with $50\%$ training epochs. Code is available at \url{https://github.com/FengLi-ust/DN-DETR}.

Code Implementations17 repos

Data from Papers with Code (CC-BY-SA-4.0)

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes