Dementia Detection on ADReSSo (test)
90.14Speaker-level F1 ScoreQwen3-14B
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-14BBackbone=Qwen3-14B, Adaptation=multi-view LoRA2026.06 | 90.14 | |
| Multimodal frameworkType=Multimodal2026.06 | 90.14 | |
| Multimodal frameworkType=Multimodal2026.06 | 89.47 | |
| Qwen3-8BBackbone=Qwen3-8B, Adaptation=multi-view LoRA2026.06 | 88.73 | |
| Gemma3-12B-itBackbone=Gemma3-12B-it, Adaptation=multi-view LoRA2026.06 | 88.72 | |
| Park et al. (2025)Type=Linguistic2026.06 | 87.5 | |
| Swin-BERTSystem=Swin-BERT2026.06 | 87.32 | |
| Qwen3-4BBackbone=Qwen3-4B, Adaptation=multi-view LoRA2026.06 | 85.66 | |
| Ilias et al. (2022)Type=Multimodal2026.06 | 85.48 | |
| Li & Zhang (2024)Type=Acoustic2026.06 | 84.51 | |
| Whisper-basedSystem=Whisper-based2026.06 | 84.5 | |
| WavBERTSystem=WavBERT2026.06 | 83.1 | |
| Multimodal frameworkType=Acoustic only2026.06 | 83.08 | |
| Zhu et al. (2021)Type=Multimodal2026.06 | 83.02 | |
| Challenge baselineSystem=Challenge baseline2026.06 | 78.92 | |
| Luz et al. (2021)Type=Multimodal2026.06 | 78.87 | |
| Multimodal frameworkType=Linguistic only2026.06 | 76.06 | |
| Luz et al. (2020)Type=Multimodal2026.06 | 75 |