论文标题
基于插值的半监督学习以进行对象检测
Interpolation-based semi-supervised learning for object detection
论文作者
论文摘要
尽管对象检测任务的数据标记成本大大远远超过分类任务，但半监督对象检测的学习方法尚未得到太多研究。在本文中，我们提出了一种基于插值的半监督学习方法，用于对象检测（ISD），该方法考虑并解决了直接将常规插值正则化（IR）直接应用于对象检测的问题。我们将模型的输出根据IR中混合的两个原始补丁的对象分数将模型的输出划分为两种类型。然后，我们以无监督的方式施加适合每种类型的单独损失。拟议的损失极大地改善了半监督学习和监督学习的表现。在监督的学习环境中，我们的方法将基线方法提高了大幅度。在半监督的学习环境中，我们的算法在基准体系结构（SSD）中提高了基准数据集（Pascal VOC和MSCOCO）的性能。
Despite the data labeling cost for the object detection tasks being substantially more than that of the classification tasks, semi-supervised learning methods for object detection have not been studied much. In this paper, we propose an Interpolation-based Semi-supervised learning method for object Detection (ISD), which considers and solves the problems caused by applying conventional Interpolation Regularization (IR) directly to object detection. We divide the output of the model into two types according to the objectness scores of both original patches that are mixed in IR. Then, we apply a separate loss suitable for each type in an unsupervised manner. The proposed losses dramatically improve the performance of semi-supervised learning as well as supervised learning. In the supervised learning setting, our method improves the baseline methods by a significant margin. In the semi-supervised learning setting, our algorithm improves the performance on a benchmark dataset (PASCAL VOC and MSCOCO) in a benchmark architecture (SSD).
