Junjie (Jorji) Chen

I was born in November 2002 in Lu’an, Anhui, China. I am a PhD student at Hefei University of Technology (HFUT), supervised by Prof. Meng Wang (汪萌). I am also working with Prof. Linfeng Zhang (张林峰) and his research group, where I am learning and exploring research problems related to Efficient AI and Foundation Models.

💡 I am broadly interested in understanding how AI systems can better model human psychological and emotional states, and how such understanding can be incorporated into practical and efficient interactive systems.

🎯 Research Focus

I am primarily interested in the following research question:

How can we build psychologically grounded, fine-grained, multimodal, efficient, evolving, and full-duplex conversational AI systems that better understand and respond to human emotions?

In particular, I am learning and exploring:

🧠 Psychology-informed modeling of emotion and cognition in dialogue systems.
🎭 Fine-grained emotion perception, including subtle, dynamic, and context-dependent affect.
🔊🖼️ Multimodal interaction, combining language with visual, acoustic, and behavioral cues.
⚡ Efficiency-aware methods for building deployable and scalable dialogue models.

For a deeper dive into my current directions and methodology, read Research Interests.

Experience

drag or scroll

Jul. 2025 - Present

Research Intern

Institute of Artificial Intelligence (IAI), Hefei Comprehensive National Science Center

Mentored by: Meng Wang

Pursuing ongoing research on emotion-aware conversational AI, emphasizing psychological grounding, multimodal affect understanding, and efficient full-duplex dialogue systems.

Mar. 2025 - July. 2025

Research Intern

TeleAI, China Telecom

Mentored by: Xi Qiu

Focused on text-to-3D human motion techniques to support controllable video generation.

Jul. 2024 - Present

Research Assistant

EPIC Lab, Shanghai Jiao Tong University

Mentored by: Linfeng Zhang

Focused on Efficient AI, with emphasis on Model Merging, KV Cache Compression, and Diffusion Cache techniques.

Apr. 2024 - Jul. 2024

Remote Intern

Lehigh University

Mentored by: Lichao Sun

Engaged in weekly remote group meetings focused on learning and academic discussion.

Jul. 2021 - Apr. 2024

Engineering Intern

Megvii Research

Mentored by: Xi Qiu

High-precision map (HDMap) data format and algorithm design; development of a visualization platform; and development of the mapping module for autonomous driving software.

News

Sep 28, 2026	🚀 Our paper “EvolvingAvatar: Interactive 3D Head Generation That Adapts as Conversations Unfold” is now available on arXiv, with an interactive project page.
Dec 17, 2025	🚀 Our paper “Towards Seamless Interaction: Causal Turn-Level Modeling of Interactive 3D Conversational Head Dynamics” (TIMAR) is now available on arXiv, with the source code released.
Nov 06, 2025	🥰 Our paper “InterCLIP-MEP: Interactive CLIP and Memory-Enhanced Predictor for Multi-modal Sarcasm Detection” has been accepted for publication in ACM Transactions on Multimedia Computing, Communications and Applications!
Sep 05, 2025	🎉 Our paper “Seeing Sarcasm Through Different Eyes: Analyzing Multimodal Sarcasm Perception in Large Vision-Language Models” has been accepted for publication in IEEE Transactions on Computational Social Systems!
Jul 03, 2025	✨ I joined the Institute of Artificial Intelligence, Hefei Comprehensive National Science Center as a research intern to study conversational AI systems grounded in psychological theories with an emphasis on efficiency.

Latest Posts

Mar 15, 2026	Research Interests

Selected Publications

EvolvingAvatar

EvolvingAvatar: Interactive 3D Head Generation That Adapts as Conversations Unfold

Junjie Chen , Fei Wang , , Yanbin Hao , Linfeng Zhang , and Meng Wang

arXiv preprint arXiv:2609.35616, 2026

arXiv Bib PDF Website

@article{chen2026evolvingavatar,
  title = {EvolvingAvatar: Interactive 3D Head Generation That Adapts as Conversations Unfold},
  author = {Chen, Junjie and Wang, Fei and Li, Kun and Nie, Yiqi and Yang, Xun and Hao, Yanbin and Zhang, Linfeng and Wang, Meng},
  journal = {arXiv preprint arXiv:2609.35616},
  year = {2026},
  url = {https://arxiv.org/abs/2609.35616},
}

TIMAR

Towards Seamless Interaction: Causal Turn-Level Modeling of Interactive 3D Conversational Head Dynamics

Junjie Chen , Fei Wang , , Dan Guo , Linfeng Zhang , and Xun Yang

arXiv preprint arXiv:2512.15340, 2025

arXiv Bib PDF Code

@article{chen2025timar,
  title = {Towards Seamless Interaction: Causal Turn-Level Modeling of Interactive 3D Conversational Head Dynamics},
  author = {Chen, Junjie and Wang, Fei and Huang, Zhihao and Zhou, Qing and Li, Kun and Guo, Dan and Zhang, Linfeng and Yang, Xun},
  journal = {arXiv preprint arXiv:2512.15340},
  url = {https://arxiv.org/abs/2512.15340},
  year = {2025},
}

DELTA

DELTA: Deliberative Multi-Agent Reasoning with Reinforcement Learning for Multimodal Psychological Counseling

Jiangnan Yang , Junjie Chen , , Yuxin Liu , Zhangling Duan , and Jie Chen

arXiv preprint arXiv:2602.04112, 2026

arXiv Bib PDF

@article{yang2026delta,
  title = {DELTA: Deliberative Multi-Agent Reasoning with Reinforcement Learning for Multimodal Psychological Counseling},
  author = {Yang, Jiangnan and Chen, Junjie and Wang, Fei and Nie, Yiqi and Liu, Yuxin and Duan, Zhangling and Chen, Jie},
  journal = {arXiv preprint arXiv:2602.04112},
  url = {https://arxiv.org/abs/2602.04112},
  year = {2026}
}

InterCLIP-MEP

InterCLIP-MEP: Interactive CLIP and Memory-Enhanced Predictor for Multi-modal Sarcasm Detection

Junjie Chen , Hang Yu , Subin Huang , Sanmin Liu and Linfeng Zhang

ACM Transactions on Multimedia Computing, Communications and Applications, 2025

DOI arXiv Bib PDF Code

@article{chen2025interclipmep,
  title = {InterCLIP-MEP: Interactive CLIP and Memory-Enhanced Predictor for Multi-modal Sarcasm Detection},
  author = {Chen, Junjie and Yu, Hang and Huang, Subin and Liu, Sanmin and Zhang, Linfeng},
  year = {2025},
  publisher = {Association for Computing Machinery},
  address = {New York, NY, USA},
  issn = {1551-6857},
  url = {https://doi.org/10.1145/3776561},
  doi = {10.1145/3776561},
  journal = {ACM Transactions on Multimedia Computing, Communications and Applications},
}

SarcasmAnalysis

Seeing Sarcasm Through Different Eyes: Analyzing Multimodal Sarcasm Perception in Large Vision-Language Models

Junjie Chen , Xuyang Liu , Subin Huang , Linfeng Zhang and Hang Yu

IEEE Transactions on Computational Social Systems, 2025

DOI arXiv Bib PDF Code

@article{chen2025seeingsarcasmdifferenteyes,
  title = {Seeing Sarcasm Through Different Eyes: Analyzing Multimodal Sarcasm Perception in Large Vision-Language Models},
  author = {Chen, Junjie and Liu, Xuyang and Huang, Subin and Zhang, Linfeng and Yu, Hang},
  year = {2025},
  journal = {IEEE Transactions on Computational Social Systems},
  url = {https://doi.org/10.1109/TCSS.2025.3608484},
  doi = {10.1109/TCSS.2025.3608484},
}

DE-ESD

DE-ESD: Dual encoder-based entity synonym discovery using pre-trained contextual embeddings

Subin Huang , Junjie Chen , , Daoyu Li , Qing Zhou , and Sanmin Liu

Expert Systems with Applications, 2025

DOI Bib PDF Code

@article{chen2025deesd,
  title = {DE-ESD: Dual encoder-based entity synonym discovery using pre-trained contextual embeddings},
  journal = {Expert Systems with Applications},
  volume = {276},
  pages = {127102},
  year = {2025},
  issn = {0957-4174},
  doi = {10.1016/j.eswa.2025.127102},
  url = {https://www.sciencedirect.com/science/article/pii/S0957417425007249},
  author = {Huang, Subin and Chen, Junjie and Yu, Chengzhen and Li, Daoyu and Zhou, Qing and Liu, Sanmin},
  keywords = {Entity synonym set, Entity synonym discovery, Dual encoder, Pre-trained language model, Contextual embedding},
}