Hello! I am a final year Ph.D. student in the Institute for Robotics and Intelligent Machines, Georgia Institute of Technology, advised by Prof. Patrico A. Vela.
In the summers of both 2024 and 2025, I was a Research Intern at Microsoft Mixed Reality, collaborating with Ben Lundell and Harpreet Sawhney. In summer 2023, I was a Applied Scientist intern at Amazon Robotics, working with Sisir Karumanchi and Shuai Han.
My research interests are in computer vision, language processing, and their integration to advance robotic intelligence. Specifically, my work spans Robotic Grasping (both 6-DoF and planar), Language Command Understanding, and addressing Open World challenges. More recently, I have been developing algorithms that leverage Large Language Models and Vision-Language Models to enable generalizable planning and spatial understanding.
You can find my resume here (updated May 2025).
Update: I am actively seeking full-time positions in industry! I am happy to connect regarding potential opportunities!
🔥 News
- 2024.05: 🎉Excited to join the Microsoft Mixed Reality team as a research scientist intern.
- 2023.07: đź“ťOne paper accepted to ICCV 2023, on a feature-based image Out-of-Distribution detection method.
- 2023.06: đź“ťOne paper accepted to IROS 2023,on an improved keypoint-based 6-DoF grasp synthesis strategy.
- 2023.05: 🎉Excited to join the Amazon Robotics stow perception team as an applied scientist intern.
- 2023.01: đź“ťOne paper accepted to ICLR 2023, on action sequence planning with the transformer model.
- 2023.01: đź“ťOne paper accepted to ICRA 2023, on Keypoint-based 6-DoF grasp detection.
- 2021.01: đź“ťTwo papers accepted to ICRA 2021, on language-conditioned robotic grasping and semantic-based pixel feature learning for the camera relocalization.
đź“– Educations
- 2021.01 - 2025 (Expected): Ph.D. in Electrical and Computer Engineering, Georgia Tech. Advised by Dr. Patricio A. Vela. Atlanta, GA, United States.
- 2019.08 - 2020.12: M.S. in Electrical and Computer Engineering, Georgia Tech. Atlanta, GA, United States.
- 2015.09 - 2019.06: B.E. in Aerospace Engineering, Beihang University. Beijing, China.
đź’» Industrial Experience
- 2025.05 - 2025.08: Research Scientist Internship, Microsoft Mixed Reality.
- Mentor: Ben Lundell;
- Topic:Inspecting the vision-action alignment in Vision-Language-Action (VLA) models.
- Redmond, WA, United States.
- 2024.05 - 2024.08: Research Scientist Internship, Microsoft Mixed Reality.
- Mentor: Ben Lundell; Co-Mentor:Harpreet Sawhney
- Topic: Reasoning on scene graphs with Large Language Models (LLMs).
- Redmond, WA, United States.
- 2023.05 - 2023.08: Applied Scientist Internship, Amazon Robotics.
- Manager: Sisir Karumanchi; Mentor:Shuai Han
- Topic: Uncertainty estimation on deep vision models for quantifying the robotic action reliability.
- Seattle, WA, United States.
đź“ť Publications
(* denotes equal contribution)
Schema-Guided Scene-Graph Reasoning based on Multi-Agent Large Language Model System (In submission)
Yiye Chen, Harpreet Sawhney, Nicholas Gyde, Yanan Jian, Jack Saunders, Patricio A. Vela, Benjamin Lundell
Code(Coming Soon) | Project(Coming Soon)
- A schema-guided multiagent LLMs framework for iterative reasoning and planning on scene graphs.
WDiscOOD: Out-of-Distribution Detection via Whitened Linear Discriminant Analysis
Yiye Chen, Yunzhi Lin, Ruinian Xu, Patricio A. Vela
- A visual representation analysis approach to identify when a deep learning model doesn’t know in the open-world setting.
- Showing effectiveness in various vision backbones, including ResNet, Vision Transformer, and CLIP vision encoder.
Planning with Language Models through Iterative Energy Minimization
Hongyi Chen*, Yilun Du*, Yiye Chen*, Patricio A. Vela, Joshua B. Tenenbaum
- An energy-based learning and interative sampling method for action sequence planning with Transformer model.
KGNv2: Separating Scale and Pose Prediction for Keypoint-based 6-DoF Grasp Synthesis on RGB-D input
Yiye Chen; Ruinian Xu; Yunzhi Lin; Hongyi Chen; Patricio A. Vela
Code | Presentation | Poster | Supplementary
- Enhances Keypoint-GraspNet (see below) by addressing scale-related issues, where scale refers to the distance of a pose towards the single-view camera.
Keypoint-GraspNet: Keypoint-based 6-DoF Grasp Generation from the Monocular RGB-D input
Yiye Chen; Yunzhi Lin; Ruinian Xu; Patricio A. Vela
Code | Presentation | Poster | Supplementary
- A keypoint-based approach for generating 6-DoF grasp poses from single-view RGB-D input.
A Joint Network for Grasp Detection Conditioned on Natural Language Commands
Yiye Chen; Ruinian Xu; Yunzhi Lin; Patricio A. Vela
- A language-conditioned robotic grasping method by fusing the visual and language embeddings.
- GASP: Gaussian Avatars with Synthetic Priors, Jack Saunders, Charlie Hewitt, Yanan Jian, Marek Kowalski, Tadas Baltrusaitis, Yiye Chen, Darren Cosker, Virginia Estellers, Nicholas Gydé, Vinay Namboodiri, Benjamin Lundell, CVPR 2025.
- Simultaneous Multi-Level Descriptor Learning and Semantic Segmentation for Domain-Specific Relocalization, Xiaolong Wu*, Yiye Chen*, CĂ©dric Pradalier, Patricio A. Vela, ICRA 2021.
đź“ś Academic Services
- Conference Reviewer: IROS’23-24, ICRA’24, CVPR’24-25, ICLR’25
- Journal Reviewer: The International Journal of Robotics Research (IJRR), IEEE Robotics and Automation Letters (RA-L), IEEE Transactions on Industrial Electronics (TIE)