arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2304.04742cs.CV
收藏

具有稳定匹配的检测Transformer

Detection Transformer with Stable Matching

Tsinghua University(清华大学);International Digital Economy Academy (IDEA)(国际数字经济研究院);Alibaba Group(阿里巴巴集团);The Hong Kong University of Science and Technology(香港科技大学);South China University of Technology(华南理工大学)

Shilong Liu, Tianhe Ren, Jiayu Chen, Zhaoyang Zeng, Hao Zhang, Feng Li, Hongyang Li, Jun Huang, Hang Su, Jun Zhu, Lei Zhang

AI总结:

本文针对DETR中不同解码器层匹配不稳定的问题,提出仅用位置度量监督正样本分类分数的原则,并设计位置监督损失和位置调制代价两种改进,在多个DETR变体及COCO基准上取得一致提升并刷新纪录。

摘要:

本文关注DEtection TRansformers(DETR)中不同解码器层之间的匹配稳定性问题。我们指出,DETR中的不稳定匹配是由多优化路径问题引起的,而DETR中的一对一匹配设计进一步凸显了这一问题。为了解决该问题,我们表明最重要的设计是使用且仅使用位置度量(如IOU)来监督正样本的分类分数。基于这一原则,我们提出了两种简单而有效的修改,将位置度量集成到DETR的分类损失和匹配代价中,分别命名为位置监督损失(position-supervised loss)和位置调制代价(position-modulated cost)。我们在多个DETR变体上验证了我们的方法。我们的方法相比基线表现出一致的改进。通过将我们的方法与DINO结合,在COCO检测基准上使用ResNet-50骨干网络,在12轮和24轮训练设置下分别达到50.4 AP和51.5 AP,在相同设置下创造了新纪录。我们使用Swin-Large骨干网络在COCO检测test-dev上达到63.8 AP。我们的代码将在https://github.com/IDEA-Research/Stable-DINO上公开。

英文摘要:

This paper is concerned with the matching stability problem across different decoder layers in DEtection TRansformers (DETR). We point out that the unstable matching in DETR is caused by a multi-optimization path problem, which is highlighted by the one-to-one matching design in DETR. To address this problem, we show that the most important design is to use and only use positional metrics (like IOU) to supervise classification scores of positive examples. Under the principle, we propose two simple yet effective modifications by integrating positional metrics to DETR's classification loss and matching cost, named position-supervised loss and position-modulated cost. We verify our methods on several DETR variants. Our methods show consistent improvements over baselines. By integrating our methods with DINO, we achieve 50.4 and 51.5 AP on the COCO detection benchmark using ResNet-50 backbones under 12 epochs and 24 epochs training settings, achieving a new record under the same setting. We achieve 63.8 AP on COCO detection test-dev with a Swin-Large backbone. Our code will be made available at https://github.com/IDEA-Research/Stable-DINO.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑