[Re]不要根据上下文判断对象:学习克服上下文偏见
[Re] Don't Judge an Object by Its Context: Learning to Overcome Contextual Bias
摘要:
Singh等人(2020)指出了视觉识别数据集中上下文偏见的危险。他们提出了两种方法,基于CAM和特征分割,可以在缺乏典型上下文的情况下更好地识别对象或属性,同时保持具有竞争力的上下文内准确性。为了验证它们的性能,我们试图复制原始论文中的所有12个表,包括附录中的表。我们还进行了额外的实验,以更好地理解所提出的方法,包括增加基于CAM的正则化和消除特征分割中的加权损失。由于原始代码不可用,我们在PyTorch 1.7.0中从头开始实现了整个管道。我们的实现是基于与作者的论文和电子邮件交流。我们发现,原始论文中提出的两种方法都有助于减轻上下文偏差,尽管对于某些方法,即使在完成广泛的超参数搜索后,我们也无法完全复制论文中的定量结果。例如,在COCO-Stuff、DeepFashion和UnRel上,我们的特征分割模型在标准基线上提高了上下文外图像的准确性,而在AwA上,我们看到了性能的下降。对于建议的CAM为基础的方法,我们能够重现原始文件的结果在0.5 $\%$ mAP。我们的实现可以在https://github.com/princetonvisualai/ContextualBias上找到。
英文摘要:
Singh et al. (2020) point out the dangers of contextual bias in visual recognition datasets. They propose two methods, CAM-based and feature-split, that better recognize an object or attribute in the absence of its typical context while maintaining competitive within-context accuracy. To verify their performance, we attempted to reproduce all 12 tables in the original paper, including those in the appendix. We also conducted additional experiments to better understand the proposed methods, including increasing the regularization in CAM-based and removing the weighted loss in feature-split. As the original code was not made available, we implemented the entire pipeline from scratch in PyTorch 1.7.0. Our implementation is based on the paper and email exchanges with the authors. We found that both proposed methods in the original paper help mitigate contextual bias, although for some methods, we could not completely replicate the quantitative results in the paper even after completing an extensive hyperparameter search. For example, on COCO-Stuff, DeepFashion, and UnRel, our feature-split model achieved an increase in accuracy on out-of-context images over the standard baseline, whereas on AwA, we saw a drop in performance. For the proposed CAM-based method, we were able to reproduce the original paper's results to within 0.5$\%$ mAP. Our implementation can be found at https://github.com/princetonvisualai/ContextualBias.
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。