On Model Calibration for Long-Tailed Object Detection and Instance Segmentation-猫眼课题宝

DOI：

发表时间：

2021-07

期刊：

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

Tai-Yu Pan;Cheng Zhang;Yandong Li;Hexiang Hu;D. Xuan;Soravit Changpinyo;Boqing Gong;Wei-Lun Chao

用于目标检测和实例分割的普通模型在长尾环境下严重偏向于检测频繁目标。现有方法主要在训练期间解决该问题，例如通过重新采样或重新加权。在本文中，我们研究了一个很大程度上被忽视的方法--置信度分数的后处理校准。我们提出了用于长尾目标检测和实例分割的Norcal，归一化校准，这是一种简单明了的方法，根据每一类的训练样本大小来重新加权每一类的预测分数。我们表明，单独处理背景类和归一化每个建议的类上分数是获得优异性能的关键。在LVIS数据集上，NorCal可以有效地改进几乎所有的基线模型，不仅在稀有类上，而且在常见类和频繁类上也是如此。最后，我们进行了广泛的分析和消融研究，以提供对我们方法的各种建模选择和机制的见解。我们的代码在https://github.com/tydpan/NorCal/.上公开提供

Vanilla models for object detection and instance segmentation suffer from the heavy bias toward detecting frequent objects in the long-tailed setting. Existing methods address this issue mostly during training, e.g., by re-sampling or re-weighting. In this paper, we investigate a largely overlooked approach -- post-processing calibration of confidence scores. We propose NorCal, Normalized Calibration for long-tailed object detection and instance segmentation, a simple and straightforward recipe that reweighs the predicted scores of each class by its training sample size. We show that separately handling the background class and normalizing the scores over classes for each proposal are keys to achieving superior performance. On the LVIS dataset, NorCal can effectively improve nearly all the baseline models not only on rare classes but also on common and frequent classes. Finally, we conduct extensive analysis and ablation studies to offer insights into various modeling choices and mechanisms of our approach. Our code is publicly available at https://github.com/tydpan/NorCal/.

On Model Calibration for Long-Tailed Object Detection and Instance Segmentation

文献摘要

相似文献