English

LightMem-Ego: Your AI Memory for Everyday Life

Computation and Language 2026-07-13 v1 Artificial Intelligence Computer Vision and Pattern Recognition Human-Computer Interaction Multimedia

Abstract

Personal AI assistants on mobile and wearable devices continuously perceive users' daily lives through visual and audio streams. However, answering queries about past experiences requires lightweight multimodal memory that can continuously accumulate, organize, and retrieve long-term experiences, which remains challenging. To address this challenge, we present LightMem-Ego, a lightweight streaming multimodal memory system for everyday-life assistance. The system continuously captures egocentric visual and audio streams, aligns them on a shared timeline, and organizes them into a hierarchical memory consisting of current, short-term, and long-term memory. Given a user query, LightMem-Ego dynamically routes retrieval to the appropriate memory level and generates answers grounded in multimodal evidence. The demonstration can be deployed on smartphones and AI glasses, supporting object finding, conversation recall, life summarization, routine discovery, and personalized assistance. Code is available at https://github.com/zjunlp/LightMem-Ego.

Keywords

Cite

@article{arxiv.2607.11487,
  title  = {LightMem-Ego: Your AI Memory for Everyday Life},
  author = {Yijun Chen and Boyi Xiao and Yixian Zhao and Haoting Xia and Buqiang Xu and Jizhan Fang and Yanya Li and Yaqi Zheng and Xuehai Wang and Zirui Xue and Liuxin Zhang and Hui Li and Ningyu Zhang},
  journal= {arXiv preprint arXiv:2607.11487},
  year   = {2026}
}

Comments

Ongoing work