### Function Introduction
AI video stream processing capabilities primarily provide two core function modules: intelligent subject enhancement and intelligent audio processing.
### Intelligent Subject Enhancement
- **Function Description**: Supports intelligent recognition and visual enhancement of specific subjects such as pets and portraits.
- **Interaction Method**:
- Users can independently select the subject category they wish to highlight.
- The system automatically completes target tracking and visual optimization processing.
### Intelligent Audio Processing
- **Mixing Control**:
- Visual interface for volume adjustment
- Adjustable original sound/background music ratio (0-100% linear adjustment)
- **Output Options**:
- Retain original sound
- Background music only
- Mixed audio output
### Interaction Process
```mermaid
sequenceDiagram
rect rgb(191, 223, 255)
User->>Panel: 1. Upload video and set parameters (subject / background music / mix ratio)
Panel->>App: 2. Initiate AI processing request (objectDetectForVideo + onVideoObjectDetectProgress)
App->>App: 3. Process video stream (calculate progress in real time)
App-->>Panel: 4. Return processing progress percentage
Panel->>Panel: Display current progress
App-->>Panel: 5. Return processed video URL
Panel->>App: 6. Unregister listener (offVideoObjectDetectProgress)
Panel->>App: Fetch first frame (fetchVideoThumbnails)
App-->>Panel: Return first frame image
Panel->>Panel: Display result video and thumbnail
end
rect rgb(200, 150, 255)
Note right of Panel: User Interaction Layer
Note left of App: AI Video Processing Service
end
```
### Considerations
- 1. Currently, subject selection only supports two categories: pets and portraits.
- 2. Mixing ratio range: 0%-100% (default is 50% for original video sound and 50% for background music).
- 3. **Code Practice** Please refer to [Video Stream Subject Highlighting Template Details](https://developer.tuya.com/en/miniapp-codelabs/codelabs/panel-ai-video-highlight/index.html#0).