7 papers
To Memories and Beyond: From Remembering to Knowing You across Long-Term Multimodal Personal Archives
Wenqi Zhou, Zhuorui Yu, Kaiao Wen +6
As AI systems evolve into personalized digital companions, a central capability is reasoning over a user's long-term personal history: not merely storing past events, but tracking…
ICM-Bench: Person-Level Identity Reasoning in Multimodal Agents with Long-Term Memory
Shidu Ren, Yunze Liu, Xing Liu +3
Long-horizon multimodal agents should remember not only what happened but also who participated. This capability depends on linking recurring faces, voices, names, person-associate…
OmniRetriever: Any-to-Any Audio-Video-Text Retrieval via Fusion-as-Teacher Distillation
Yunze Liu, Chi-Hao Wu, Enmin Zhou +1
Unified multimodal embedding spaces have become the standard interface for cross-modal retrieval and multimodal RAG, and recent audio-video-text (AVT) encoders extend this setting…
O-MARC: Omni Memory-Augmented Compression Distillation for Efficient Video Understanding
Peiran Wu, Yunze Liu, Chi-Hao Wu +2
Omnimodal large language models enable unified audio video understanding, but long joint token sequences make inference costly, and existing benchmarks do not fully isolate audio v…
Bridging Modalities, Spanning Time: Structured Memory for Ultra-Long Agentic Video Reasoning
Jiazheng Li, Chi-Hao Wu, Yunze Liu +3
Understanding ultra-long videos such as egocentric recordings, live streams, or surveillance footage spanning days to weeks, remains a challenge. For current multimodal LLMs: even…
SpatialMem: Metric-Aligned Long-Horizon Video Memory for Language Grounding and QA
Xinyi Zheng, Yunze Liu, Chi-Hao Wu +5
SpatialMem is a memory-based system which integrates 3D geometry, semantics, and language with one another in one combined, queryable representation. Starting with casually recorde…
MARC: Memory-Augmented RL Token Compression for Efficient Video Understanding
Peiran Wu, Zhuorui Yu, Yunze Liu +3
The rapid progress of large language models (LLMs) has laid the foundation for multimodal models. However, visual language models (VLMs) still face heavy computational costs when e…