Multimodal Representation Learning Conditioned on Semantic Relations

arXiv:2508.17497v3 Announce Type: replace-cross Abstract: Multimodal representation learning has been largely driven by contrastive models such as CLIP, which learn a shared embedding space by aligning paired image-text samples. While effective for general-purpose representation learning, such models typically produce a single embedding per sample that is reused across different semantic relations and contexts. However, in many real-world applications, relevance between samples is inherently relation-dependent, with different semantic relations emphasizing different aspects of multimodal data. In this work, we propose Relation-Conditioned Multimodal Learning (RCML), a framework that treats semantic relations as explicit conditions of multimodal representation learning. Rather than producing relation-agnostic embeddings, RCML learns representations conditioned on natural-language relation descriptions, allowing the same sample to be represented differently under different relational contexts. The framework constructs relation-aware training pairs, introduces a relation-conditioned module to adapt embeddings to relation semantics, and employs a unified contrastive objective to jointly model cross-modal alignment and relation-induced inter-sample structure. Experiments on multiple datasets show that RCML consistently outperforms strong baselines on retrieval and classification tasks in zero-shot, fine-tuned, and out-of-domain settings, highlighting the effectiveness of leveraging semantic relations to guide multimodal representation learning.

Abstract:Multimodal representation learning has been largely driven by contrastive models such as CLIP, which learn a shared embedding space by aligning paired image-text samples. While effective for general-purpose representation learning, such models typically produce a single embedding per sample that is reused across different semantic relations and contexts. However, in many real-world applications, relevance between samples is inherently relation-dependent, with different semantic relations emphasizing different aspects of multimodal data. In this work, we propose Relation-Conditioned Multimodal Learning (RCML), a framework that treats semantic relations as explicit conditions of multimodal representation learning. Rather than producing relation-agnostic embeddings, RCML learns representations conditioned on natural-language relation descriptions, allowing the same sample to be represented differently under different relational contexts. The framework constructs relation-aware training pairs, introduces a relation-conditioned module to adapt embeddings to relation semantics, and employs a unified contrastive objective to jointly model cross-modal alignment and relation-induced inter-sample structure. Experiments on multiple datasets show that RCML consistently outperforms strong baselines on retrieval and classification tasks in zero-shot, fine-tuned, and out-of-domain settings, highlighting the effectiveness of leveraging semantic relations to guide multimodal representation learning.

Machine Learning (cs.LG); Artificial Intelligence (cs.AI)

arXiv:2508.17497 [cs.LG]

(or arXiv:2508.17497v3 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2508.17497 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

aiqbench analysis

Multimodal Representation Learning Conditioned on Semantic Relations

Related coverage