List your company

Paralinguistic Understanding on Paralinguistic Understanding Tasks (test)

61.1Emotion AccuracyKimi-Audio-7B
21.47631.76342.0552.337May 7, 2026

Evaluation results

MethodLinks
Kimi-Audio-7BBackbone=7B2026.0561.175.960.531.257.2
TextPro-SLM-7BBackbone=7B2026.0560.588.664.945.164.8
TextPro-SLM-3BBackbone=3B2026.0558.880.263.74561.9
TextPro-SLM-7B 5:1Backbone=7B, Prosody Embedding Ratio=5:12026.0557.788.866.750.265.8
TextPro-SLM-7B w/o ReconBackbone=7B, Ablation variant=Without reconstruction objective2026.0556.779.964.244.461.3
Qwen2.5-Omni-7BBackbone=7B2026.0554.889.844.845.358.7
TextPro-SLM-7B w/o TrainBackbone=7B, Ablation variant=Without paralinguistic training2026.0551.828.543.616.135
Qwen2-Audio-7BBackbone=7B2026.0548.264.723.12941.3
DiVA2026.0530.746.833.332.535.8
GLM-4-Voice-9BBackbone=9B2026.052323.918.732.424.5