Hybrid Batch Training for Multilingual CLIP in Zero-Shot Image-Text Retrieval

Information retrieval across different languages is an increasingly important challenge in natural language processing. Recent approaches based on multilingual pre-trained language models have achieved remarkable success, yet they often optimize for either monolingual, cross-lingual, or multilingual retrieval performance at the expense of others. This paper proposes a novel hybrid batch training strategy to simultaneously improve zero-shot retrieval performance across monolingual, cross-lingual, and multilingual settings while mitigating language bias. The approach fine-tunes multilingual lang

Research goal: Does the hybrid batch training strategy proposed in this paper improve zero-shot image-text retrieval accuracy on the XTD benchmark when applied to multilingual variants of CLIP, compared to standard monolingual or cross-lingual fine-tuning approaches?

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