Using Multiple Instance Learning to Build Multimodal Representations
This work addresses the challenge of building effective multimodal representations for medical applications like image classification and cross-modal retrieval, offering a generic framework that improves performance in specific tasks.
The paper tackled the problem of aligning image-text multimodal representations for medical applications by connecting multimodal representation learning with multiple instance learning, resulting in a novel contrastive learning approach that achieved state-of-the-art results in downstream tasks.
Image-text multimodal representation learning aligns data across modalities and enables important medical applications, e.g., image classification, visual grounding, and cross-modal retrieval. In this work, we establish a connection between multimodal representation learning and multiple instance learning. Based on this connection, we propose a generic framework for constructing permutation-invariant score functions with many existing multimodal representation learning approaches as special cases. Furthermore, we use the framework to derive a novel contrastive learning approach and demonstrate that our method achieves state-of-the-art results in several downstream tasks.