About me

I'm Poppy (Yingjing) Xu, currently an AI Researcher at ModelBest（面壁智能）, where my work focuses on post-training of Omni Streaming-based Multimodal Large Language Model (MLLM).

I received my Master's degree in Software Engineering from Zhejiang University, co-advised by Dr. Bo Lin and Prof. Jianwei Yin at the ACES-HiG Lab.

What I'm interested in

  1. Omni Model Training. A solid data-construction pipeline for omni-modal models, and training strategies that get closer to the essence of omni intelligence.
  2. Omni Evaluation. Benchmarks that reflect a model's real foundational ability, and that work as a guideline for building higher-value intelligence.
  3. Real-world Omni Agents. Omni & voice agents that make a positive influence in daily life.

Feel free to get in touch via Email. :)

News

  1. Feb, 2026

    Our omni foundation model, MiniCPM-o-4.5 released, the world’s first omnimodal full-duplex model.

  2. Jan, 2025

    Joined ModelBest（面壁智能）, building MiniCPM V and o series as a team member, focusing on post-training of multi-modal large language models.

  3. June, 2024

    Joined 01.AI (零一万物), working on instruction-based image editing.

  4. June, 2022

    Joined Binjiang Institute of Zhejiang University, focusing on human-centered medical AI and computer vision.

Selected Work

  • MiniCPM-o 4.5

    MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction

    Core Contributor — omni simplex & duplex post-training

    arXiv, 2026

    Paper Project

  • MiniCPM-V 4.5

    MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe

    Contributor — multilingual OCR post-training

    CVPR, 2026

    Paper Project

  • Omni-DuplexEval

    Omni-DuplexEval: Evaluating Real-time Duplex Omni-modal Interaction

    Core Contributor

    arXiv, 2026

    Paper Project

  • Liberating LLM Capabilities

    Liberating LLM Capabilities in Full-Duplex Speech Models

    Contributor

    COLM, 2026

    Paper Project

  • InsightEdit

    InsightEdit: Towards Better Instruction Following for Image Editing

    First Author — instruction-following image editing

    CVPR, 2025

    Paper Project

Daniel lewis
quote icon

Daniel lewis

Richard was hired to create a corporate identity. We were very pleased with the work done. She has a lot of experience and is very concerned about the needs of client. Lorem ipsum dolor sit amet, ullamcous cididt consectetur adipiscing elit, seds do et eiusmod tempor incididunt ut laborels dolore magnarels alia.

Publications

  • MiniCPM-o 4.5

    MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction

    Junbo Cui, Bokai Xu, Chongyi Wang, Tianyu Yu, Weiyue Sun, Yingjing Xu, Tianran Wang, et al.

    arXiv, 2026

    Paper Project

  • MiniCPM-V 4.5

    MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe

    Tianyu Yu, Zefan Wang, Chongyi Wang, ..., Yingjing Xu, ..., Yuan Yao, Zhiyuan Liu, Maosong Sun

    CVPR, 2026

    Paper Project

  • Omni-DuplexEval

    Omni-DuplexEval: Evaluating Real-time Duplex Omni-modal Interaction

    Chaoqun He*, Mingyang Xiang*, Yingjing Xu, Bokai Xu, Junbo Cui, Jie Zhou, Yuan Yao, Lijie Wen

    arXiv, 2026

    Paper Project

  • Liberating LLM Capabilities

    Liberating LLM Capabilities in Full-Duplex Speech Models

    Luoyuan Zhang, Bokai Xu, Junbo Cui, Weiyue Sun, Yingjing Xu, Hanyu Liu, Yuan Yao

    COLM, 2026

    Paper Project

  • InsightEdit

    InsightEdit: Towards Better Instruction Following for Image Editing

    Yingjing Xu*, Jie Kong*, Jiazhi Wang, Xiao Pan, Bo Lin, and Qiang Liu

    CVPR, 2025

    Paper Project

  • SkinGEN

    SkinGEN: An Explainable Dermatology Diagnosis-to-generation Framework with Interactive Vision-Language Models

    Bo Lin*, Yingjing Xu*, Xuanwen Bao, Zhou Zhao, Zhouyang Wang, Jianwei Yin

    IUI, 2025

    Paper

  • CECT-Mamba

    CECT-Mamba: a Hierarchical Contrast-enhanced-aware Model for Pancreatic Tumor Subtyping from Multi-phase CECT

    Zhifang Gong, Shuo Gao, Ben Zhao, Yingjing Xu, Yijun Yang, Shenghong Ju, Guangquan Zhou

    ICCV Workshops, 2025

    Paper

  • Break the Mask Barrier

    "Break the Mask Barrier": An AU-based Rehabilitation Training System for Parkinson's Hypomimia

    Xueyan Cai*, Yingjing Xu*, Zihong Zhou, Mengru Xue, Zhengke Li, Chentian Weng, Wei Luo, Cheng Yao, Bo Lin, Jianwei Yin

    CHI EA, 2025

    Paper

  • HypomimiaCoach

    HypomimiaCoach: An AU-based Digital Therapy System for Hypomimia Detection & Rehabilitation with Parkinson's Disease

    Yingjing Xu*, Xueyan Cai*, Zihong Zhou, Mengru Xue, Bo Wang, Haotian Wang, Zhengke Li, Chentian Weng, Wei Luo, Cheng Yao, Bo Lin, Jianwei Yin

    arXiv, 2024

    Paper

  • Facial AU-aid hypomimia diagnosis

    Facial AU-aid hypomimia diagnosis based on GNN

    Yingjing Xu, Bo Lin, Wei Luo, Shuiguang Deng, and Jianwei Yin

    MIDL, 2023

    Paper

Projects