Jiahao Nick LI 李嘉昊

I am a Ph.D. candidate at UCLA HCI Research working with Prof. Xiang ‘Anthony’ Chen.

My research area lies in Human-AI Interaction, focusing on Egocentric Multimodal AI Agents. Specifically, I build interactive systems to provide personalized assistance for individual users. Such systems are powered by cutting-edge AI models (e.g., LLMs, VLMs) and are enhanced by incorporating: (1) external and task-specific knowledge, sourced through both crowdsourcing and targeted data collection, and (2) the retrieval of long-form multimodal memories, facilitating contextual understanding for users. Recently, I have been working on building a multimodal retrieval augmented pipeline to address the natural language query (NLQ) task on the egocentric video datasets (e.g., Ego4D).

During my Ph.D., I interned at Meta Reality Lab, Adobe Research and Palo Alto Research Center (PARC). I was also a visiting Ph.D. student at University of Tokyo working with Dr. Takeo Igarashi. Prior to that, I received my Bachelor of Engineering from Shanghai Jiao Tong University.

news

Jan 19, 2024 Two papers (OmniActions & Human I/O) got accepted by CHI 2024 :tada:. OmniActions is from my Meta internship on user intent prediction in pervasive AR scenarios leveraging LLMs and VLMs. Human I/O is on identifying situational impairments using egocentric videos leveraging reasonings of LLMs.

selected publications

  1. omniactions.gif
    OmniActions: Predicting Digital Actions in Response to Real-World Multimodal Sensory Inputs with LLMs
    Jiahao Nick Li, Yan Xu, Tovi Grossman, and 2 more authors
    In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems (CHI ’24), 2024
  2. humanio_thumbnail.gif
    Human I/O: Towards a Unified Approach to Detecting Situational Impairments in Everyday Activities
    Xingyu Bruce Liu, Jiahao Nick Li, David Kim, and 2 more authors
    In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems (CHI ’24), 2024
  3. roman.gif
    Roman: Making Everyday Objects Robotically Manipulable with 3D-Printable Add-on Mechanisms
    Jiahao Li, Alexis Samoylov, Jeeeun Kim, and 1 more author
    In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems (CHI ’22), 2022
  4. romeo.gif
    Romeo: A Design Tool for Embedding Transformable Parts in 3D Models to Robotically Augment Default Functionalities
    Jiahao Li, Meilin Cui, Jeeeun Kim, and 1 more author
    In Proceedings of the 33rd Annual ACM Symposium on User Interface Software and Technolog (UIST ’20), 2020
  5. robiot.gif
    Robiot: A Design Tool for Actuating Everyday Objects with Automatically Generated 3D Printable Mechanisms
    Jiahao Li, Jeeeun Kim, and Xiang ’Anthony’ Chen
    In Proceedings of the 32nd Annual ACM Symposium on User Interface Software and Technology (UIST ’19), 2019