List your company

Multimodal Machine Translation on TopicVD (test)

30.47BLEUProposed Method
18.686821.745924.80527.8641Apr 8, 2026

Evaluation results

MethodLinks
Proposed MethodRetrieval size P=10, Context window w=2, Scaling factor gamma=0.1, Selection count K=5, Fusion weight lambda=0.12026.0430.4738.96
Segment-Level Video-MMTInput=temporally aligned video segment, Context=local2026.0429.2338.65
Image-MMTVisual features=extracted from sampled frames, Retrieval=FAISS2026.0428.9338.59
BigVideoLearning=cross-modal contrastive learning2026.0425.9236.54
Text-only NMTArchitecture=Transformer-based2026.0419.1433.95