List your company

Speech-to-text translation on English-to-Chinese (En->Zh)

49.1BLEU ScoreVIRBA
40.46842.70944.9547.191Aug 25, 2025

Evaluation results

MethodLinks
VIRBABackbone=Qwen2.5-Omni2025.0849.1
VIRBABackbone=Qwen2-Audio2025.0848.4
Kimi-Audio2025.0848.2
Qwen2.5-Omni2025.0848
Group-DPO2025.0847.3
Step-Audio-R12025.0847.1
AF32025.0846.8
SeamlessM4T-L V22025.0846.2
Qwen2-Audiovariant=base2025.0845.2
Whisper-large V22025.0840.8