SOTAVerified

Word Alignment

Word Alignment is the task of finding the correspondence between source and target words in a pair of sentences that are translations of each other.

Source: Neural Network-based Word Alignment through Score Aggregation

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–10 of 551 papers

TitleStatusHype
The Devil Is in the Word Alignment Details: On Translation-Based Cross-Lingual Transfer for Token Classification TasksCode0
Large Language Models Often Say One Thing and Do AnotherCode0
Target word activity detector: An approach to obtain ASR word boundaries without lexicon—0
BinaryAlign: Word Alignment as Binary Sequence LabelingCode1
Cross-Lingual Word Alignment for ASEAN Languages with Contrastive Learning—0
Tradition or Innovation: A Comparison of Modern ASR Methods for Forced Alignment—0
Learning Language Structures through Grounding—0
Word Alignment as Preference for Machine Translation—0
Image-Text Co-Decomposition for Text-Supervised Semantic SegmentationCode1
Enhancing Cross-lingual Sentence Embedding for Low-resource Languages with Word Alignment—0
Show:102550
← PrevPage 1 of 56Next →
All datasetsen-esen-fres-enfr-enen-itMUSE en-deMUSE en-pt

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1Barycenter AlignmentP@184.26—Unverified
2Adv - Refine - CSLSP@181.7—Unverified