Contextual Classification Using Self-Supervised Auxiliary Models for Deep Neural Networks
This work addresses the problem of improving classification performance and interpretability for deep neural networks by integrating contextual signals, which is beneficial for researchers and practitioners developing more robust and understandable AI systems.
This paper introduces Self-Supervised Autogenous Learning (SSAL) models to incorporate contextual signals into deep neural networks for classification. SSAL models achieve faster convergence and consistently outperform state-of-the-art methods while providing more interpretable structured predictions.
Classification problems solved with deep neural networks (DNNs) typically rely on a closed world paradigm, and optimize over a single objective (e.g., minimization of the cross-entropy loss). This setup dismisses all kinds of supporting signals that can be used to reinforce the existence or absence of a particular pattern. The increasing need for models that are interpretable by design makes the inclusion of said contextual signals a crucial necessity. To this end, we introduce the notion of Self-Supervised Autogenous Learning (SSAL) models. A SSAL objective is realized through one or more additional targets that are derived from the original supervised classification task, following architectural principles found in multi-task learning. SSAL branches impose low-level priors into the optimization process (e.g., grouping). The ability of using SSAL branches during inference, allow models to converge faster, focusing on a richer set of class-relevant features. We show that SSAL models consistently outperform the state-of-the-art while also providing structured predictions that are more interpretable.