脸谱心智
基础模型与学习范式
脸谱心智开发端侧多模态模型，并以“叠叠社”等实时 AI 弹幕产品探索交互方式。系统可理解屏幕上的文字、图像与声音，生成与当前内容相关的回应。
FaceMind develops on-device multimodal models and tests them in real-time products such as DieDieShe. Its AI commentary can respond to text, images and audio on a user's screen, creating a more context-aware interaction.
