Skip to content

Latest commit

 

History

History
# README

Use speaker 1 for debug.

## 1. Prepare dataset

### Audio and gesture

### Text

We only use self-talk dataset, witch text is available now.

Transcript 83 is lost.

Install Gentle. (Follow Trimodal)

## 2. KNN based matching


## 3. Codebook

## 4. DeepPhase



## 5.