Scaling Multilingual Multimodal Embeddings for Zero-Shot Text-to-Video Retrieval Performance

This paper studies zero-shot cross-lingual transfer of vision-language models. Specifically, we focus on multilingual text-to-video search and propose a Transformer-based model that learns contextualized multilingual multimodal embeddings. Under a zero-shot setting, we empirically demonstrate that performance degrades significantly when we query the multilingual text-video model with non-English sentences. To address this problem, we introduce a multilingual multimodal pre-training strategy, and collect a new multilingual instructional video dataset (MultiHowTo100M) for pre-training. Experimen

Research goal: Does scaling the size of multilingual multimodal embeddings correlate with improved performance on non-English queries in zero-shot text-to-video retrieval benchmarks?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.8/10.