cv
Basics
| Name | Tewodros Wondifraw Ayalew |
| Label | PhD Student |
| tewodrosayalew [at] uchicago [dot] edu | |
| Url | https://github.com/tedywond |
| Summary | PhD student in Computer Science at the University of Chicago with experience in computer vision, machine learning, and plant phenotyping. |
Education
-
2022.09 - Present Chicago, Illinois
-
2018.09 - 2020.11 Saskatoon, Saskatchewan
-
2011.09 - 2016.07 Addis Ababa, Ethiopia
Work
-
2021.10 - Present Machine Learning Engineer
Global Institute for Food Security
Developing computer vision methods for phenotyping plant roots to extract information predictive of yield.
-
2021.01 - 2021.09 Machine Learning Engineer
University of Saskatchewan and Honda R&D
Worked on improving semantic segmentation models for autonomous navigation under adverse weather using unsupervised domain adaptation.
Publications
-
2026.08 CLAW: Learning Continuous Latent Action World Models via Adversarial Latent Regularization
WM@Booth 2026 (Lightning Talk); also MWRW 2026
End-to-end continuous latent actions + diffusion world model from action-free video; planning/imitation on sim and real UR5.
-
2026.06 What Are We Actually Benchmarking in Robot Manipulation?
CoRL 2026 (Oral); IROS 2026 RGMCW Workshop (Best Abstract Award)
Four diagnostics showing popular manipulation benchmarks often fail as proxies for general capability.
-
2025.10 PROGRESSOR: A Perceptually Guided Reward Estimator with Self-Supervised Online Refinement
ICCV 2025
Self-supervised visual progress reward with online push-back; Meta-World RL and real-UR5 few-shot offline RL under noisy demos.
-
2025.03 -
2020.08 Earlier computer vision work in plant phenotyping
ECCV Workshop (CVPPP) 2020; Plant Phenomics 2021
ECCV-W CVPPP 2020 Best Paper, Plant Phenomics 2021, AutoCount.
Awards
- 2026
Best Abstract Award
IROS 2026 Workshop on Reproducible Benchmarking of Robotic Grasping and Manipulation
Awarded for 'What Are We Actually Benchmarking in Robot Manipulation?'
- 2025
CIFAR Inclusive AI Scholarship
CIFAR Deep Learning and Reinforcement Learning Summer School
- 2022
Crerar Fellowship
University of Chicago
- 2020
Best Paper Award
Awarded for 'Unsupervised Domain Adaptation for Plant Organ Counting' at CVPPP Workshop, ECCV 2020.
Volunteer
-
2018.06 - Present Programming Instructor
AddisCoder
Designed and taught a summer course in Scratch and Python for high school students.