Learning Aligned Cross-modal Representations From Weakly Aligned Data

Abstract

People can recognize scenes across many different modalities beyond natural images. In this paper, we investigate how to learn cross-modal scene representations that transfer across modalities. To study this problem, we introduce a new cross-modal scene dataset. While convolutional neural networks can categorize cross-modal scenes well, they also learn an intermediate representation not aligned across modalities, which is undesirable for cross-modal transfer applications. We present methods to regularize cross-modal convolutional neural networks so that they have a shared representation that is agnostic of the modality. Our experiments suggest that our scene representation can help transfer representations across modalities for retrieval. Moreover, our visualizations suggest that units emerge in the shared representation that tend to activate on consistent concepts independently of the modality.

Stats

• citations98
• S2 citations—
• github stars0
• HF likes0
• heat score14.97
• arxiv keycastrejon2016learning

Related papers

• Cross-modal Scene Networks (2016)14.35
• See, Hear, And Read: Deep Aligned Representations (2017)0.00
• Crossover: 3D Scene Cross-modal Alignment (2025)4.52
• Adversarial Cross-modal Retrieval Via Learning And Transferring Single-modal Similarities (2019)8.60
• Multimodal Representation Alignment For Cross-modal Information Retrieval (2025)0.00
• Discriminative Semantic Transitive Consistency For Cross-modal Learning (2021)0.00
• Preserving Semantic Neighborhoods For Robust Cross-modal Retrieval (2020)10.07
• Robust Cross-modal Representation Learning With Progressive Self-distillation (2022)12.33