About

I am Yi Ge (Ellen), a fervent researcher with a deep passion for broad area of computer vision, machine learning and robotics. During my graduate in CMU, I used to work in Robotics Institute under Prof. Sebastian Scherer and Prof. Peng Yin, and Electrical and Computer Engineering Department under Prof. Zhiqiang Shen.


Contact

Email: ellen530yige[at]gmail[dot]com
[Google Scholar] |  [Github]

Research Interests

My research interests lie at the intersection of Vision and Language, with a focus on developing efficient and robust Multimodal Systems by combining Large Language Models (LLMs), Vision-Language Models (VLVMs), and Data Embodied AI. Recently, I mainly tackle challenges in

  • Synthetic Annotated Data Generation and Compositionality (e.g., AIGC, Layout Diffusion, Dataset Diffusion).
  • Domain Adaptation, Meta Learning and Transferability (e.g., UDA, partial UDA).
  • Large-scale Language or Visual-language model for Improved Reasoning.
  • Visual SLAM, Multi-view Stereo, 3D Semantic Segmentation.

News


Selected Publications

[Peer-reviewed Conference & Journal & Workshop]
(* = equal contribution; My scholar name: Ellen Yi-Ge, Yi Ge)

FlexDataset: Crafting Annotated Dataset Generation for Diverse Applications
Ellen Yi-Ge, Leo Shawn
AAAI 2025, Poster. Accepted. 39th AAAI Conference on Artificial Intelligence.
Paper  |  Code  |  Video  |  Poster
A framework for generating high-fidelity synthetic and annotated datasets using a composition-to-data (C2D) paradigm.
Reducing Divergence in Batch Normalization for Domain Adaptation
Ellen Yi-Ge, Mingjing Wu, Zhenghan Chen
AAAI 2025, Oral. Accepted. 39th AAAI Conference on Artificial Intelligence.
Paper  |  Code
A normalization paradigm for unsupervised domain adaptation that mitigates estimation shifts accumulated in batch normalization layers.
Adversarial Domain Adaptation via Real Data Augmentation and Uncertainty Penalization
Zhenghan Chen(*), Yi Ge(*), Guohua Dong, Lei Zhu
TPAMI 2024. Under Minor Revision. IEEE Transactions on Pattern Analysis and Machine Intelligence.
Paper
Tackles partial domain adaptation challenges by leveraging augmented source data and a penalty mechanism for uncertain predictions.
Sharpness-Aware Minimization for Adversarial Domain Adaptation
Zhenghan Chen(*), Yi Ge(*)
TNNLS 2024. Under Revision. IEEE Transactions on Neural Networks and Learning Systems.
Paper
A framework for adversarial domain adaptation that integrates Interchangeable Batch Normalization (InterBN) and sharpness-aware optimization.
SphereVLAD++: Attention-based and Signal-enhanced Viewpoint Invariant Descriptor
Shiqi Zhao, Peng Yin, Yi Ge, Sebastian Scherer
RA-L 2022. Published. IEEE Robotics and Automation Letters.
Paper  |  BibTex  |  Code
An advanced method for 3D place recognition using LiDAR data.
Conditional Link Prediction of Category-Implicit Keypoint Detection
Ellen Yi-Ge, Rui Fan, Zechun Liu, Zhiqiang Shen
WACV 2021. Published. 26th Winter Conference on Applications of Computer Vision.
Paper  |  BibTex  |  Talk
A network for semantic keypoint detection and link prediction in multi-class instances.
Learning 3D Segmentation from Sparse Annotations via Hierarchical Descriptors
Ellen Yi-Ge, Rui Fan, Zechun Liu, Zhiqiang Shen
ITCA 2021. Published. 2nd International Conference on Information Technology and Computer Application.
Paper  |  Demo  |  Mini Demo
A approach that can simultaneously learn segmentation from sparse annotations via reasoning global-regional structures and individual-vicinal properties.
Importance-Aware Semantic Segmentation in Self-Driving with Discrete Wasserstein Training
Xiaofeng Liu, Yuzhuo Han, Yi Ge, Song Bai, Tianxing Wang, Xu Han, Site Li, Jane You, Jun Lu
AAAI 2020, Oral. Published. 34th AAAI Conference on Artificial Intelligence.
Paper  |  BibTex
Incorporates the importance-aware inter-class correlation in a Wasserstein training framework by configuring its ground distance matrix.
Unimodal-uniform Constrained Wasserstein Training for Medical Diagnosis
Xiaofeng Liu, Xu Han, Yukai Qiao (*), Yi Ge (*), Lu Jun
ICCVW 2019, Oral. Published. International Conference on Computer Vision Workshops.
Paper  |  BibTex
Addresses the challenges associated with discrete and successively distributed labels in medical diagnosis, like diabetic retinopathy, by incorporating the inter-class correlations in the loss function.

[Preprints & Other Projects]

DRUM: Learning Demonstration Retriever for Large MUlti-modal Models
Ellen Yi-Ge, Jiechao Gao, Wei Han, Wei Zhu
Paper (arXiv)
Demonstration retrieval tailored for large vision-language models (LVLMs).
Topic: LLM-Adaptive Low-rank Adaptation
Ellen Yi-Ge, Wei Zhu
Paper (arXiv)
Aligns the time series tokens to the language modality by attending to text prompts’ embeddings.
[Project] Parallel Density-Based Spatial Clustering
Ellen Yi-Ge
Project Website
Optimizes the DBSCAN algorithm for large datasets using various parallel computing strategies.
Tools: OpenMP, SIMD, CUDA, MPI, Cython.
[Project] High Dynamic Range Imaging Pipeline
Ellen Yi-Ge
Project Website
HDR imaging pipeline integrating image merging, white balance correction, and tone mapping

Awards and Honors

  • Graduation with Distinction, CMU
  • Dean’s Honor List for Academic Excellence, CMU
  • Kaggle Competition Master Tier - Silver Medal (Top 5%)
  • World Mind Olympics - Top 3
  • National Undergraduate Academic Competition - Golden Price
  • International Mathematical Contest in Modeling - Meritorious Winner
  • SJTU merit-based Scholarship
  • SJTU Outstanding Volunteer Service Award
  • SJTU Social Practice Award