ASLP-lab commited on
Commit
cffb5e1
·
verified ·
1 Parent(s): 4aaa430

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -5
README.md CHANGED
@@ -7,11 +7,7 @@ pipeline_tag: audio-classification
7
  ---
8
  Leveraging the multi-dimensional fine-grained annotations produced by our pipeline, we introduce **FM-Speech**, built upon the frontier Qwen3-Omni (30B MoE) architecture.
9
 
10
- <div align="center">
11
- <!-- 请在此处放一张精简的模型输入输出示意图 -->
12
- <!--<img width="500px" src="src/model_io.png" alt="Model Input Output Diagram"/> -->
13
- <p><i>Input: Raw Speech &emsp; ➔ &emsp; Output: 14-Dimension Fine-Grained Speech Attributes (Structured JSON)</i></p>
14
- </div>
15
 
16
  To overcome modality gaps and text-conditioned hallucinations, FM-Speech is trained using a **Progressive Curriculum Fine-Tuning** framework, decoupling complex auditory comprehension into three incremental stages: Warm-up (MCQ/QA) $\rightarrow$ Capability Ramp-up $\rightarrow$ Final Alignment (Full JSON).
17
 
 
7
  ---
8
  Leveraging the multi-dimensional fine-grained annotations produced by our pipeline, we introduce **FM-Speech**, built upon the frontier Qwen3-Omni (30B MoE) architecture.
9
 
10
+ > 🎙️ **Input:** Raw Speech Audio &emsp; ➔ &emsp; 📊 **Output:** 14-Dimension Fine-Grained Speech Attributes (Structured JSON)
 
 
 
 
11
 
12
  To overcome modality gaps and text-conditioned hallucinations, FM-Speech is trained using a **Progressive Curriculum Fine-Tuning** framework, decoupling complex auditory comprehension into three incremental stages: Warm-up (MCQ/QA) $\rightarrow$ Capability Ramp-up $\rightarrow$ Final Alignment (Full JSON).
13