Autoencoder-based Feature Extraction for the Classification Tasks with Insufficient Dataset Size
Emil Guliyev, Nurhan Aliyev, Nihad Asgarov, Jamaladdin Hasanov · 2023
The presented work proposes an effective approach for extracting abstract characteristics from image data using the autoencoder-based models. Since simple autoencoders do not deliver the desired result in building a feature map between the data samples, variations and domain-specific adjustments might improve the performance. To assist a model with more informative and representative samples, augmentation technique on small subset with the position and size invariance have been applied. To evaluate the efficiency, we employ simple autoencoder and U-Net models that take both data features and their relationships into consideration. The suggested autoencoder models are assessed on a collection of benchmark datasets, and the experimental findings demonstrate that, in comparison to other autoencoder variants, taking data relationships into account can lead to more robust features that accomplish reduce construction loss and then reduced rate of errors in subsequent classification.