问题描述
我正在使用 Tensorflow 的 对象检测 API,但得到以下内容训练时出错:
I'm using Tensorflow's Object Detection API, but get the following error when training:
InvalidArgumentError(回溯见上文):断言失败:[最大框坐标值大于1.01:] [1.47]
当我使用以下任何一项时出现错误:
I get the error when I use any of the following:
- faster_rcnn_inception_resnet_v2_atrous_coco
- rfcn_resnet101_coco
但当我使用时不会:
- ssd_inception_v2_coco
- ssd_mobilenet_v1_coco
我的训练图像混合了 300x300 和 450x450 像素.我不相信我的任何边界框都在图像坐标之外.即使是这种情况,为什么最后两个模型可以工作,而 resnet 模型却不能?
My training images are a mixture of 300x300 and 450x450 pixels. I don't believe any of my bounding boxes are outside the image coordinates. Even if that's the case why would the last two models work but not the resnet models?
推荐答案
在查看我的原始边界框数据后,结果发现有一些随机实例,其中边界框坐标要么是非常大的数要么是负数(不确定那是如何开始的).我删除了这些,现在训练任何模型都没有问题.
After looking at my raw bounding box data, turns out there were a few random instances where the bounding box coordinates either had very large numbers or negative numbers (not sure how that happened to begin with). I deleted these and now I have no issue training any of the models.
这篇关于对象检测 API 断言失败:[最大框坐标值大于 1.01:] 中的 resnet 模型的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!