Teaching Compositionality to CNNs
This addresses the issue of poor generalization in CNNs for computer vision by enhancing feature compositionality, though it appears incremental as it builds on existing CNN architectures.
The paper tackles the problem of making CNNs learn compositional features that disentangle objects from their surroundings, resulting in more localized feature activations and improved performance in object recognition tasks over non-compositional baselines.
Convolutional neural networks (CNNs) have shown great success in computer vision, approaching human-level performance when trained for specific tasks via application-specific loss functions. In this paper, we propose a method for augmenting and training CNNs so that their learned features are compositional. It encourages networks to form representations that disentangle objects from their surroundings and from each other, thereby promoting better generalization. Our method is agnostic to the specific details of the underlying CNN to which it is applied and can in principle be used with any CNN. As we show in our experiments, the learned representations lead to feature activations that are more localized and improve performance over non-compositional baselines in object recognition tasks.