LightMem-Ego is a lightweight streaming multimodal memory system for everyday-life assistants running on smartphones and AI glasses. It continuously processes egocentric visual and audio streams, aligns them on a shared timeline, and organizes experiences into current-, short-term-, and long-term-memory levels. For each query, the system dynamically routes retrieval to an appropriate level and generates answers grounded in multimodal evidence. The paper describes applications including object finding, conversation recall, life summarization, routine discovery, and personalized assistance. Its implementation is available on GitHub, but the supplied abstract does not report quantitative benchmarks or resource measurements.
No heat snapshots are available in the last 24 hours.