Abstract
In this paper we investigate the problems of class imbalance and irrelevant relationships in Visual Relationship Detection (VRD). State-of-the-art deep VRD models still struggle to predict uncommon classes, limiting their applicability. Moreover, many methods are incapable of properly filtering out background relationships while predicting relevant ones. Although these problems are very apparent, they have both been overlooked so far. We analyse why this is the case and propose modifications to both model and training to alleviate the aforementioned issues, as well as suggesting new measures to complement existing ones and give a more holistic picture of the efficacy of a model.
Abstract (translated)
本文研究了视觉关系检测(VRD)中的类不平衡和无关关系问题。先进的深VRD模型仍然难以预测不常见的类,限制了它们的适用性。此外,许多方法在预测相关的背景关系时都不能正确地过滤掉背景关系。尽管这些问题很明显,但迄今为止都被忽视了。我们分析了这种情况的原因,并提出了对模型和培训的修改,以缓解上述问题,同时提出了补充现有问题的新措施,并对模型的有效性进行了更全面的描述。
URL
https://arxiv.org/abs/1903.08456