DanceBack Agent MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@DanceBack Agent MCP Servermirror my frontal dance video to practice from the back"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
DanceBack Agent
把正面舞蹈视频转成背面跟练视角,并通过 MCP / Function Calling 集成到 AI Agent。
项目目标
主线:正面视频 → 3D 姿态 → 绕 Y 轴旋转 180° → 背面数字人渲染
降级:左右镜像,快速跟练(fallback)
Related MCP server: chisel
当前版本 v0.1
项目骨架
占位 Skill:镜像(fallback)
MCP Server 骨架(适配 mcp 2.x)
2D 姿态提取(MediaPipe Tasks API)
3D 姿态提升(v0.2)
背面数字人渲染(v0.3)
快速开始
1. 安装依赖
pip install -e .2. 下载模型文件
运行姿态提取前,请先下载 MediaPipe 模型:
Invoke-WebRequest -Uri "https://storage.googleapis.com/mediapipe-models/pose_landmarker/pose_landmarker_full/float16/latest/pose_landmarker_full.task" -OutFile "pose_landmarker_full.task"3. 镜像视频(fallback 模式)
python src\danceback\skills\mirror.py dance.mp4 practice.mp44. 提取 2D 姿态
python src\danceback\skills\extract_pose.py
生成 pose_preview.jpg。5. 启动 MCP Server
python src\danceback\agent\mcp_server.py目录结构
danceback-agent/
├── src/danceback/
│ ├── skills/
│ │ ├── mirror.py # fallback:左右镜像
│ │ ├── back_view.py # 主线:背面视角(占位)
│ │ └── extract_pose.py # 2D 姿态提取
│ └── agent/
│ └── mcp_server.py # MCP Server
├── pyproject.toml
└── README.md路线图
版本 | 目标 |
v0.0.1 | 骨架 + 镜像占位 |
v0.1 | MCP Server + 2D 姿态提取 |
v0.2 | 3D 姿态提升 |
v0.3 | 背面数字人渲染 |
说明
back_view 是主线,mirror 仅作 fallback。
单目正面视频无法真实恢复背面细节,背面为推断结果。
pose_landmarker_full.task 和 pose_preview.jpg 已在 .gitignore 中忽略,不会上传。
请勿上传未授权舞蹈视频和音乐。
许可证
MIT
This server cannot be deployed
Maintenance
Related MCP Connectors
AI video, images, music & SFX: Seedance 2.5, Veo 3.1, Kling 3.0, Nano Banana Pro, 20+ models.
AI visual generation agent: multi-pipeline rendering, prompt crafting, and image composition.
Generate, edit and upscale AI video and images from any agent via VicSee.
AI influencer images and video generation and editing for your agent.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables AI agents to analyze images and videos, and generate optimized prompts for AI video generation systems.MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI agents to construct, edit, and export 3D models using geometric primitives and boolean operations, with multi-view rendering to facilitate spatial reasoning.7MIT
- AlicenseNot gradedqualityBmaintenanceEnables AI agents to control video editing software (剪映/CapCut and Adobe Premiere Pro) through a unified interface, supporting operations like material import, clip splitting, subtitle addition, effects, transitions, audio mixing, and export.10MIT
- AlicenseCqualityDmaintenanceMCP server for video editing with text-prompt subject segmentation and compositing effects behind subjects. Enables agent-driven video creation and education animations.275MIT