↓Skip to main content

Minkyoung Cho

Ph.D. Student in CSE at the University of Michigan

Photo of Minkyoung Cho

I am a final-year Ph.D. student, fortunate to be advised by Prof. Z. Morley Mao. My research broadly asks: How can we bridge the modality gap to build AI systems that truly understand and interact with the physical world?

I explore this question across multimodal and physical AI, with particular interests in aligning and transferring information across modalities and in grounding perception and reasoning into action. My recent work spans multimodal learning, efficient adaptation, entity-centric reasoning, and vision-language-action models.

I am on the job market for full-time positions as a Ph.D. new graduate, starting May 2027.

Selected Publications

I view the modality gap broadly. Different sources of information differ in their representations, reliability, dynamics, and roles, and AI systems need mechanisms to reconcile these discrepancies.

My research has explored this gap in progressively broader forms: across timesteps and sensing modalities, in conditioning and multimodal learning dynamics, and most recently, between the digital and physical worlds.

  1. BMVC 2023

    Current ↔ Future . Aspects: reliability.

    ADoPT

    Feature gap across timesteps

  2. ICLR 2025

    2D Vision ↔ 3D Vision . Aspects: representations, reliability.

    Cocoon

    Modality reliability / representation gap

  3. NeurIPSW 2025

    Condition ↔ Generation . Aspects: dynamics, roles.

    TC-LoRA

    Condition ↔ Generation dynamics gap

  4. NeurIPS 2026

    Vision ↔ Language . Aspects: dynamics.

    MARS

    Learning-dynamics gap across modalities

  5. In Submission

    Digital ↔ Physical . Aspects: representations, roles.

    GroundAct

    Digital ↔ Physical grounding gap

All Publications

  1. In Submission

    SiA: Synergy in Action: Benchmarking Collaborative Coordination in Multi-Arm Furniture Assembly

    Zesen Zhao*, Joseph Brewington*, Minkyoung Cho*, Ruiyang Zhu, Haoran Zhang, Hui Shen, Boyuan Zheng, Jiarui Li, Xueshen Liu, Fan Bai, Shuqing Zeng, Z. Morley Mao (*: equal contribution)

  2. NeurIPS 2026

    MARS: Harmonizing Multimodal Convergence via Adaptive Rank Search

    Minkyoung Cho, Insu Jang, Shuowei Jin, Zesen Zhao, Adityan Jothi, Ethem Can, Min-Hung Chen, Z. Morley Mao

    Paper Webpage
    BibTeX for MARS: Harmonizing Multimodal Convergence via Adaptive Rank Search 
    @inproceedings{cho2026mars,
  title={MARS: Harmonizing Multimodal Convergence via Adaptive Rank Search},
  author={Cho, Minkyoung and Jang, Insu and Jin, Shuowei and Zhao, Zesen and Jothi, Adityan and Can, Ethem F. and Chen, Min-Hung and Mao, Z. Morley},
  booktitle={Advances in Neural Information Processing Systems (NeurIPS)},
  year={2026}
}
  3. IROS 2026

    GraphMap: Scalable Crowd-Sourced Global Vectorized HD Map Construction via Sparse Visual Graph Fusion

    Ruiyang Zhu, Minkyoung Cho, Shuqing Zeng, Fan Bai, Z. Morley Mao

    Paper
    BibTeX for GraphMap: Scalable Crowd-Sourced Global Vectorized HD Map Construction via Sparse Visual Graph Fusion 
    @article{zhu2026graphmap,
  title={GraphMap: Scalable Crowd-Sourced Global Vectorized HD Map Construction via Sparse Visual Graph Fusion},
  author={Zhu, Ruiyang and Cho, Minkyoung and Zeng, Shuqing and Bai, Fan and Mao, Z. Morley},
  year={2026}
}
  4. ICML 2026

    Dynamic Linear Attention

    Xin Wang*, Hui Shen*, Boyuan Zheng, Xueshen Liu, Minkyoung Cho, Zhongwei Wan, Zesen Zhao, Zhuoqing Mao, Shen Yan, Mi Zhang (*: equal contribution)

    Paper
    BibTeX for Dynamic Linear Attention 
    @inproceedings{wang2026dla,
  title={{Dynamic Linear Attention}},
  author={Wang, Xin and Shen, Hui and Zheng, Boyuan and Liu, Xueshen and Cho, Minkyoung and Wan, Zhongwei and Zhao, Zesen and Mao, Zhuoqing and Yan, Shen and Zhang, Mi},
  booktitle={International Conference on Machine Learning (ICML)},
  year={2026}
}
  5. NeurIPSW 2025

    TC-LoRA: Temporally Modulated Conditional LoRA for Adaptive Diffusion Control

    Minkyoung Cho, Ruben Ohana, Christian Jacobsen, Adityan Jothi, Min-Hung Chen, Z. Morley Mao, Ethem Can

    Paper Webpage Poster
    BibTeX for TC-LoRA: Temporally Modulated Conditional LoRA for Adaptive Diffusion Control 
    @inproceedings{cho2025tclora,
  title={TC-LoRA: Temporally Modulated Conditional LoRA for Adaptive Diffusion Control},
  author={Minkyoung Cho, Ruben Ohana, Christian Jacobsen, Adityan Jothi, Min-Hung Chen, Z. Morley Mao, Ethem Can},
  journal={arXiv preprint arXiv:2510.09561},
  year={2025}
}
  6. IROS 2025 Oral

    SCORPION: Robust Spatial-Temporal Collaborative Perception Model on Lossy Wireless Network

    Ruiyang Zhu, Minkyoung Cho, Shuqing Zeng, Fan Bai, Z. Morley Mao

    Paper Webpage Poster
    BibTeX for SCORPION: Robust Spatial-Temporal Collaborative Perception Model on Lossy Wireless Network 
    @inproceedings{zhu2025scorpion,
  title={SCORPION: Robust Spatial-Temporal Collaborative Perception Model on Lossy Wireless Network},
  author={Ruiyang Zhu, Minkyoung Cho, Shuqing Zeng, Fan Bai, and Z. Morley Mao},
  journal={2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  year={2025}
}
  7. CVPRW 2025

    CrowdMap: Scalable Crowd-sourced Global HD Map Construction via Collaborative Map Perception and Sparse Graph Fusion

    Ruiyang Zhu*, Minkyoung Cho*, Shuqing Zeng, Fan Bai, Xiang Gao, Z. Morley Mao (*: equal contribution)

    Paper Poster
    BibTeX for CrowdMap: Scalable Crowd-sourced Global HD Map Construction via Collaborative Map Perception and Sparse Graph Fusion 
    @inproceedings{zhu2025crowdmap,
  title={Scalable Crowd-sourced Global HD Map Construction via Collaborative Map Perception and Sparse Graph Fusion},
  author={Ruiyang Zhu, Minkyoung Cho, Shuqing Zeng, Fan Bai, Xiang Gao, and Z. Morley Mao},
  journal={CVPR Workshop 2025},
  year={2025}
}
  8. ICLR 2025

    Cocoon: Robust Multi-Modal Perception with Uncertainty-Aware Sensor Fusion

    Minkyoung Cho, Yulong Cao, Jiachen Sun, Qingzhao Zhang, Marco Pavone, Jeong Joon Park, Heng Yang, Z. Morley Mao

    Paper Webpage Poster
    BibTeX for Cocoon: Robust Multi-Modal Perception with Uncertainty-Aware Sensor Fusion 
    @inproceedings{cho2025cocoon,
  title={Cocoon: Robust Multi-Modal Perception with Uncertainty-Aware Sensor Fusion},
  author={Cho, Minkyoung and Cao, Yulong and Sun, Jiachen and Zhang, Qingzhao and Pavone, Marco and Park, Jeong Joon and Yang, Heng and Mao, Zhuoqing},
  booktitle={The Thirteenth International Conference on Learning Representations}
}
  9. CSCS 2024

    Achieving the Safety and Security of the End-to-End AV Pipeline

    Noah T. Curran, Minkyoung Cho, Ryan Feng, Liangkai Liu, Brian Jay Tang, Pedram MohajerAnsari, Alkim Domeke, Mert D. Pesé, Kang G. Shin  ACM CSCS Workshop Co-located with ACM CCS 2024

    Paper
    BibTeX for Achieving the Safety and Security of the End-to-End AV Pipeline 
    @inproceedings{curran2023achieving,
  title={Achieving the safety and security of the end-to-end av pipeline},
  author={Curran, Noah T and Cho, Minkyoung and Feng, Ryan and Liu, Liangkai and Tang, Brian Jay and MohajerAnsari, Pedram and Domeke, Alkim and Pes{\'e}, Mert D and Shin, Kang G},
  booktitle={Proceedings of the 2024 Cyber Security in CarS Workshop},
  pages={13--24},
  year={2023}
}
  10. BMVC 2023

    ADoPT: LiDAR Spoofing Attack Detection based on Point-Level Temporal Consistency

    Minkyoung Cho, Yulong Cao, Zixiang Zhou, Z. Morley Mao

    Paper Poster
    BibTeX for ADoPT: LiDAR Spoofing Attack Detection based on Point-Level Temporal Consistency 
    @inproceedings{cho2023adopt,
  author    = {Minkyoung Cho and Yulong Cao and Zixiang Zhou and Zhuoqing Morley Mao},
  title     = {ADoPT: LiDAR Spoofing Attack Detection Based on Point-Level Temporal Consistency},
  booktitle = {34th British Machine Vision Conference 2023, {BMVC} 2023, Aberdeen, UK, November 20-24, 2023},
  publisher = {BMVA},
  year      = {2023},
  url       = {https://papers.bmvc2023.org/0486.pdf}
}
  11. arXiv 2023

    DynaMIX: Resource Optimization for DNN-Based Real-Time Applications on a Multi-Tasking System

    Minkyoung Cho and Kang G. Shin

    Paper
    BibTeX for DynaMIX: Resource Optimization for DNN-Based Real-Time Applications on a Multi-Tasking System 
    @article{cho2023dynamix,
  title={Dynamix: Resource optimization for dnn-based real-time applications on a multi-tasking system},
  author={Cho, Minkyoung and Shin, Kang G},
  journal={arXiv preprint arXiv:2302.01568},
  year={2023}
}
  12. ICIP 2021

    A Novel Sensitivity Metric For Mixed-Precision Quantization With Synthetic Data Generation

    Donghyun Lee*, Minkyoung Cho*, Seungwon Lee, Joonho Song, Changkyu Choi (*: equal contribution)

    Paper
    BibTeX for A Novel Sensitivity Metric For Mixed-Precision Quantization With Synthetic Data Generation 
    @inproceedings{lee2021novel,
  title={A novel sensitivity metric for mixed-precision quantization with synthetic data generation},
  author={Lee, Donghyun and Cho, Minkyoung and Lee, Seungwon and Song, Joonho and Choi, Changkyu},
  booktitle={2021 IEEE International Conference on Image Processing (ICIP)},
  pages={1294--1298},
  year={2021},
  organization={IEEE}
}
  13. M4IoT 2016

    Contextual Relationship-based Activity Segmentation on an Event Stream in the IoT Environment with Multi-user Activities

    Minkyoung Cho, Younggi Kim, Younghee Lee  ACM M4IoT Workshop Co-located with Middleware 2016

    Paper
    BibTeX for Contextual Relationship-based Activity Segmentation on an Event Stream in the IoT Environment with Multi-user Activities 
    @article{cho2016contextual,
  title={Contextual relationship-based activity segmentation on an event stream in the iot environment with multi-user activities},
  author={Cho, Minkyoung and Kim, Younggi and Lee, Younghee},
  journal={arXiv preprint arXiv:1609.06024},
  year={2016}
}
  14. INFOCOM 2016

    Proactive patrol dispatch surveillance system by inferring mobile trajectories of multiple intruders using binary proximity sensors

    Dahee Jung, Minkyoung Cho, Omprakash Gnawali, HyungJune Lee

    Paper
    BibTeX for Proactive patrol dispatch surveillance system by inferring mobile trajectories of multiple intruders using binary proximity sensors 
    @inproceedings{jeong2016proactive,
  title={Proactive patrol dispatch surveillance system by inferring mobile trajectories of multiple intruders using binary proximity sensors},
  author={Jeong, Dahee and Cho, Minkyoung and Gnawali, Omprakash and Lee, HyungJune},
  booktitle={IEEE INFOCOM 2016-The 35th Annual IEEE International Conference on Computer Communications},
  pages={1--9},
  year={2016},
  organization={IEEE}
}

Patents

Research Experience

Academia

  • 2022 – Present

    UMich  Graduate Student Research Assistant : Physical AI

  • 2021 – 2022

    UMich  Research Intern : Model Optimization

  • 2015 – 2017

    KAIST  Research Assistant

Industry

  • 2026

    NVIDIA Research  ASPIRE  Ph.D. Research Intern : VLA Reasoning

  • 2025

    NVIDIA  Metropolis  Deep Learning SW Intern : Conditional Diffusion

  • 2024

    NVIDIA  Metropolis  Deep Learning SW Intern : PEFT

  • 2018 – 2021

    Samsung Advanced Institute of Technology (SAIT)  AI Researcher : Model Optimization

Education

Service

Organizing

Tutorial Organizer, X-Sense Tutorial, ECCV 2026  Sep 2026

Reading Group Organizer, Systems Reading Group, UMich CSE  Sep 2023 – Present

Reviewing
Conferences
Workshops