List your company

Speech Recognition on Vaani 48 languages (test)

27.4WERSigLIP2-large
27.372427.558727.74527.9313Jun 23, 2026

Evaluation results

MethodLinks
SigLIP2-largeinput_mode=multi-image, alignment=Audio-Image Alignment2026.0627.40.00692.470.0062
Qwen3-VLinput_mode=multi-image, alignment=Audio-Image Alignment2026.0627.680.00411.470.0033
SigLIP2-baseinput_mode=single-image, alignment=Audio-Image Alignment2026.0627.710.00381.350.003
Baselinealignment=no alignment2026.0628.09———