Papers with MERLIN
MERLIN: A Testbed for Multilingual Multimodal Entity Recognition and Linking (2026.tacl-1)
Copied to clipboard
Sathyanarayanan Ramamoorthy, Vishwa Shah, Simran Khanuja, Zaid Sheikh, Shan Jie, Ann Chia, Shearman Chua, Graham Neubig
| Challenge: | Existing methods for multilingual entity linking are limited by textual contexts and limited resources. |
| Approach: | They propose a testbed system for multilingual multimodal entity linking using BBC news articles paired with corresponding images in five languages. |
| Outcome: | The proposed system improves accuracy for entities with ambiguous textual contexts and models with weak multilingual abilities. |
MERLIN: Multimodal Embedding Refinement via LLM-based Iterative Navigation for Text-Video Retrieval-Rerank Pipeline (2024.emnlp-industry)
Copied to clipboard
| Challenge: | Recent advances in text-video retrieval neglect the crucial user perspective, leading to discrepancies between user queries and content retrieved. |
| Approach: | They propose a novel, training-free pipeline that leverages Large Language Models for iterative feedback learning. |
| Outcome: | Experimental results show that MERLIN significantly outperforms existing systems in video retrieval. |
Automatic Extraction of Nominal Phrases from German Learner Texts of Different Proficiency Levels (2024.lrec-main)
Copied to clipboard
| Challenge: | a pilot study has found that inflecting determiners and adjectives correctly is a challenge for learners of German. |
| Approach: | They propose to use dependency parsing to extract nouns, grammatical heads and dependents that have to agree with the noun in German. |
| Outcome: | The proposed method performs well on CEFR levels A1-B1 but not level B2 texts. |
MERLIN: Multi-Stage Curriculum Alignment for Multilingual Encoder-LLM Integration in Cross-Lingual Reasoning (2026.eacl-long)
Copied to clipboard
Kosei Uemura, David Guzmán, Quang Phuoc Nguyen, Jesujoba Oluwadara Alabi, En-Shiun Annie Lee, David Ifeoluwa Adelani
| Challenge: | Existing methods to align large language models with multilingual encoders raise accuracy for low-resource languages (LRLs) but performance of LLMs in low- and high-resourced languages remains a problem. |
| Approach: | They propose a model-stacking framework that iteratively refines in 2-stages based on a curriculum strategy and adapts only a small set of DoRA weights. |
| Outcome: | The proposed framework improves exact-match accuracy by +12.9 pp over MindMerger and outperforms GPT-4o-mini by 15.2 pp on the AfriMGSM benchmark. |