LightMem-Ego: Lightweight Streaming Multimodal Memory for Smartphones and AI Glasses

ZJUNLP

Research official 2 src. ~1 min

A lightweight streaming multimodal memory system designed for smartphones and AI glasses that continuously processes visual and audio streams and organizes them into a three-tier hierarchy (current, short-term, long-term memory). Dynamic routing directs queries to the appropriate tier, enabling object finding, conversation recall, life summarization, and routine discovery. Code is open-sourced.

Why it matters

Bridges the gap between on-device wearable sensing and personal AI assistants; hierarchical architecture keeps inference lightweight enough for mobile hardware while preserving long-horizon recall. Received 25 upvotes on HuggingFace Daily Papers.

Importance: 2/5

HuggingFace Daily Papers (25 upvotes); practical mobile AI memory architecture for wearables and always-on agents

Sources