是否可以从Keras的flow_from_directory自动推断出class_weight? [英] Is it possible to automatically infer the class_weight from flow_from_directory in Keras?

问题描述

我有一个不平衡的多类数据集，我想使用fit_generator中的class_weight参数根据每个类的图像数量为这些类赋予权重.我正在使用ImageDataGenerator.flow_from_directory从目录中加载数据集.

I have an imbalanced multi-class dataset and I want to use the class_weight argument from fit_generator to give weights to the classes according to the number of images of each class. I'm using ImageDataGenerator.flow_from_directory to load the dataset from a directory.

是否可以直接从ImageDataGenerator对象推断出class_weight自变量?

Is it possible to directly infer the class_weight argument from the ImageDataGenerator object?

推荐答案

只是想出了一种方法来实现这一目标.

Just figured out a way of achieving this.

from collections import Counter
train_datagen = ImageDataGenerator()
train_generator = train_datagen.flow_from_directory(...)

counter = Counter(train_generator.classes)                          
max_val = float(max(counter.values()))       
class_weights = {class_id : max_val/num_images for class_id, num_images in counter.items()}                     

model.fit_generator(...,
                    class_weight=class_weights)

train_generator.classes是每个图像的类的列表. Counter(train_generator.classes)创建一个每个类别中图像数量的计数器.

train_generator.classes is a list of classes for each image. Counter(train_generator.classes) creates a counter of the number of images in each class.

请注意，这些权重可能不利于收敛，但是您可以将其用作基于发生率的其他类型加权的基础.

Note that these weights may not be good for convergence, but you can use it as a base for other type of weighting based on occurrence.

此答案的灵感来自于: https://github.com /fchollet/keras/issues/1875#issuecomment-273752868

这篇关于是否可以从Keras的flow_from_directory自动推断出class_weight?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IT屋！