COCO-DR Continuous Contrastive Learning for Cross-Domain Multimodal Retrieval on Flickr30k

Multimodal sarcasm detection aims to identify sarcasm in the given image-text pairs and has wide applications in the multimodal domains. Previous works primarily design complex network structures to fuse the image-text modality features for classification. However, such complicated structures may risk overfitting on in-domain data, reducing the performance in out-of-distribution (OOD) scenarios. Additionally, existing methods typically do not fully utilize cross-modal features, limiting their performance on in-domain datasets. Therefore, to build a more reliable multimodal sarcasm detection mo

Research goal: Does the COCO-DR approach of continuous contrastive learning on target corpora improve cross-domain generalization for multimodal retrieval on Flickr30k relative to in-distribution performance?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 8.5/10.