如何从深度 CNN 输出创建不确定性彩色图图像？

Question

我正在设计一个用于城市特征检测的深度 CNN 分类器。大多数时候我的网络正确地分类和分割建筑物，但很多时候它会因为光照/相似的外观等与其他对象混淆。

我想创建一个颜色图和分割图像来表示特定分类器的程度？我使用 softmaxwith loss 来训练网络。

layer {
  name: "score"
  type: "Deconvolution"
  bottom: "pool_3"
  top: "score"
  convolution_param {
    num_output: 2
    bias_term: false
    pad:2
    kernel_size: 8
    stride: 4
  }
}

我期望输出类似于此彩色地图图像：

我的问题是

如何计算不确定性？
计算不确定性时如何处理负值？

注意：目前，我可以使用熵得到颜色图。

Answer 1

您可能想要执行遮挡敏感度实验，以便构建最重要图像区域的热图。

来自this answer on AI StackExchange：

Here's the idea. Suppose that a ConvNet classifies an image as a dog. How can we be certain that it’s actually picking up on the dog in the image as opposed to some contextual cues from the background or some other miscellaneous object?

One way of investigating which part of the image some classification prediction is coming from is by plotting the probability of the class of interest (e.g. dog class) as a function of the position of an occluder object. If we iterate over regions of the image, replace it with all zeros and check the classification result, we can build a 2-dimensional heat map of what's most important for the network on a particular image.

如何从深度 CNN 输出创建不确定性彩色图图像？

How do I create uncertainty color map image from deep CNN output?

statistics

probability

neural-network

deep-learning

conv-neural-network