Search papers, labs, and topics across Lattice.
PersonaVLM is introduced as a framework to enable long-term personalization in MLLMs by incorporating memory, reasoning, and response alignment capabilities. The framework extracts and summarizes multimodal memories, retrieves relevant memories for multi-turn reasoning, and infers user personality to align outputs. Evaluated on a new benchmark, Persona-MME, PersonaVLM demonstrates significant improvements over baselines and even outperforms GPT-4o in long-term personalization tasks.
Forget static, single-turn personalization – PersonaVLM unlocks long-term, evolving user alignment in MLLMs, even surpassing GPT-4o.
Multimodal Large Language Models (MLLMs) serve as daily assistants for millions. However, their ability to generate responses aligned with individual preferences remains limited. Prior approaches enable only static, single-turn personalization through input augmentation or output alignment, and thus fail to capture users'evolving preferences and personality over time (see Fig.1). In this paper, we introduce PersonaVLM, an innovative personalized multimodal agent framework designed for long-term personalization. It transforms a general-purpose MLLM into a personalized assistant by integrating three key capabilities: (a) Remembering: It proactively extracts and summarizes chronological multimodal memories from interactions, consolidating them into a personalized database. (b) Reasoning: It conducts multi-turn reasoning by retrieving and integrating relevant memories from the database. (c) Response Alignment: It infers the user's evolving personality throughout long-term interactions to ensure outputs remain aligned with their unique characteristics. For evaluation, we establish Persona-MME, a comprehensive benchmark comprising over 2,000 curated interaction cases, designed to assess long-term MLLM personalization across seven key aspects and 14 fine-grained tasks. Extensive experiments validate our method's effectiveness, improving the baseline by 22.4% (Persona-MME) and 9.8% (PERSONAMEM) under a 128k context, while outperforming GPT-4o by 5.2% and 2.0%, respectively. Project page: https://PersonaVLM.github.io.
