Gaussian Mixture Variational Autoencoder with Contrastive Learning for Multi-Label Classification
This addresses multi-label classification, a common problem in machine learning applications like image tagging, by offering a data-driven method that reduces the need for complex modules, though it appears incremental in combining existing techniques.
The paper tackles multi-label classification by proposing C-GMVAE, a model that uses a Gaussian mixture variational autoencoder with contrastive learning to capture label correlations and improve prediction, achieving performance matching other models with only 50% of training data on multiple datasets.
Multi-label classification (MLC) is a prediction task where each sample can have more than one label. We propose a novel contrastive learning boosted multi-label prediction model based on a Gaussian mixture variational autoencoder (C-GMVAE), which learns a multimodal prior space and employs a contrastive loss. Many existing methods introduce extra complex neural modules like graph neural networks to capture the label correlations, in addition to the prediction modules. We find that by using contrastive learning in the supervised setting, we can exploit label information effectively in a data-driven manner, and learn meaningful feature and label embeddings which capture the label correlations and enhance the predictive power. Our method also adopts the idea of learning and aligning latent spaces for both features and labels. In contrast to previous works based on a unimodal prior, C-GMVAE imposes a Gaussian mixture structure on the latent space, to alleviate the posterior collapse and over-regularization issues. C-GMVAE outperforms existing methods on multiple public datasets and can often match other models' full performance with only 50% of the training data. Furthermore, we show that the learnt embeddings provide insights into the interpretation of label-label interactions.