Teamhelper Research Lab focuses on the frontier intersection of AI × AR × Edge Computing, with emphasis on on-device large models, spatial computing, and human–computer interaction.

We conduct systematic research on the integration of AR smart glasses and artificial intelligence, addressing core challenges including:

• intent understanding and tool planning
• multi-device collaborative computation
• on-device inference and acceleration
• large-scale system optimization

Our goal is to bring large-model applications on low-power devices from theory to real-world deployment.

• Lightweight on-device large models
  training, compression, and deployment

• Intent recognition & multi-tool planning
  interpretable frameworks and execution pipelines

• Collaborative edge computation
  LAN-based P2P compute and task-graph scheduling

• Spatial computing prototypes
  AR-centric interaction systems and experimental platforms

We collaborate with:

• university interns
• early-career researchers
• industrial partners

Together, we build an open research and engineering ecosystem delivering:

• open-source code
• models
• datasets
• technical reports

We aim to advance the next generation of intelligent AR-native devices through rigorous research and practical engineering.

Teamhelper Research Lab 聚焦 AI × AR × 边缘计算 的前沿交叉方向，重点关注 端侧大模型、空间计算与人机交互。

我们围绕 AR 眼镜与人工智能的深度融合 开展系统性研究，重点面向以下核心技术挑战：

• 意图识别与工具规划
• 多设备协同计算
• 端侧推理与加速优化
• 大规模系统工程化部署

我们的目标是推动 大模型在低功耗设备上的应用，实现从 理论方法到真实场景落地 的跨越。

• 端侧轻量化大模型
  训练、压缩与部署

• 可解释意图识别与多工具规划
  框架设计与执行链路

• 边缘协同计算
  基于局域网的 P2P 互联与任务图调度

• 空间计算原型系统
  面向 AR 的交互系统与实验平台

我们与以下伙伴协同开展研究：

• 高校实习生
• 初级研究人员
• 产业合作伙伴

共建开放的科研与工程生态，持续产出：

• 开源代码
• 模型
• 数据集
• 技术报告

我们致力于以 严谨的科研方法与工程实践 推动 新一代 AR 原生智能设备 的发展。