Fuzzy Pooling
This work addresses uncertainty in CNNs for image classification, offering a drop-in replacement for pooling layers, but it is incremental as it builds on existing pooling operators.
The paper tackles the problem of uncertainty propagation in convolutional neural networks by proposing a novel pooling operation based on fuzzy sets to handle local imprecision in feature maps, and it shows that this approach enhances classification performance and outperforms state-of-the-art pooling methods in experiments on publicly available datasets.
Convolutional Neural Networks (CNNs) are artificial learning systems typically based on two operations: convolution, which implements feature extraction through filtering, and pooling, which implements dimensionality reduction. The impact of pooling in the classification performance of the CNNs has been highlighted in several previous works, and a variety of alternative pooling operators have been proposed. However, only a few of them tackle with the uncertainty that is naturally propagated from the input layer to the feature maps of the hidden layers through convolutions. In this paper we present a novel pooling operation based on (type-1) fuzzy sets to cope with the local imprecision of the feature maps, and we investigate its performance in the context of image classification. Fuzzy pooling is performed by fuzzification, aggregation and defuzzification of feature map neighborhoods. It is used for the construction of a fuzzy pooling layer that can be applied as a drop-in replacement of the current, crisp, pooling layers of CNN architectures. Several experiments using publicly available datasets show that the proposed approach can enhance the classification performance of a CNN. A comparative evaluation shows that it outperforms state-of-the-art pooling approaches.