On Binary Classification with Single-Layer Convolutional Neural Networks
This work addresses the challenge of optimizing convolutional network design for binary classification tasks, but it is incremental as it builds on existing methods with specific improvements.
The paper tackles the problem of designing single-layer convolutional neural networks for binary classification, showing that pre-training, regularization, and pool size are critical factors, and achieves performance close to state-of-the-art SVM models on a cats and dogs dataset.
Convolutional neural networks are becoming standard tools for solving object recognition and visual tasks. However, most of the design and implementation of these complex models are based on trail-and-error. In this report, the main focus is to consider some of the important factors in designing convolutional networks to perform better. Specifically, classification with wide single-layer networks with large kernels as a general framework is considered. Particularly, we will show that pre-training using unsupervised schemes is vital, reasonable regularization is beneficial and applying of strong regularizers like dropout could be devastating. Pool size is also could be as important as learning procedure itself. In addition, it has been presented that using such a simple and relatively fast model for classifying cats and dogs, performance is close to state-of-the-art achievable by a combination of SVM models on color and texture features.