Adversarial Cross-Modal Retrieval
University of Electronic Science and Technology of China · Delft University of Technology
Abstract
Cross-modal retrieval aims to enable flexible retrieval experience across different modalities (e.g., texts vs. images). The core of cross-modal retrieval research is to learn a common subspace where the items of different modalities can be directly compared to each other. In this paper, we present a novel Adversarial Cross-Modal Retrieval (ACMR) method, which seeks an effective common subspace based on adversarial learning. Adversarial learning is implemented as an interplay between two processes. The first process, a feature projector, tries to generate a modality-invariant representation in the common subspace and to confuse the other process, modality classifier, which tries to discriminate between…
Citation impact
- FWCI
- 28.28
- Percentile
- 100%
- References
- 43
Authors
5- BWBokun WangCorresponding
University of Electronic Science and Technology of China
- YYYang Yang
University of Electronic Science and Technology of China
- XXXing Xu
University of Electronic Science and Technology of China
- AHAlan Hanjalić
Delft University of Technology
- HTHeng Tao Shen
University of Electronic Science and Technology of China
Topics & keywords
- Computer science
- Subspace topology
- Modal
- Modalities
- Artificial intelligence
- Modality (human–computer interaction)
- Classifier (UML)
- Representation (politics)
- Reduced inequalities