关注智能系统在真实约束下的行为。以下条目属于持续研究或探索，不代表完成的产品，也不以未经核验的指标作结论。I explore how intelligent systems behave under real constraints. These entries describe ongoing or exploratory work—not finished products or claims supported by unverified metrics.
探索视觉、语言与动作模型在机器人任务中的连接方式，作为 SmartCam 的研究扩展。当前关注系统接口与实验方法，不作为独立完成产品展示。Exploring connections between vision, language and action models for robotics, as a research extension of SmartCam. The focus is on interfaces and experimental methods rather than a separate finished product.
围绕 sEMG 手势识别与适应性研究，关注信号处理、实验组织和评估流程。数据许可与个体差异是验证过程中需要明确的约束。Research around sEMG gesture recognition and adaptation, with attention to signal processing, experimental organization and evaluation. Data permissions and individual variation remain explicit constraints.
通过模型复现探索分割训练与评估流程。这里仅介绍研究方向，不公开医疗影像、个人信息、受限数据或未经核验的临床结论。Exploring segmentation training and evaluation through model reproduction. This entry describes the research direction only; no medical images, personal information, restricted data or unverified clinical conclusions are published.