Yash Jangir

Yash Jangir

CS PhD Student, Johns Hopkins University

🤖 Robotics 👁️ Computer Vision 🧠 Embodied AI

Hello, a person I don't know about :)

I’m Yash, a student of Homanga Bharadhwaj. My research is motivated by a deceptively simple question: How does intelligence enable an organism to interact with the world? I study how robots can learn and reason about dynamics, both those of the physical world and those that emerge through interaction, to take actions that anticipate, plan for, and adapt under uncertainty. Or in fancier terms, I work on world models, multimodal AI, and robot learning for contact-rich and dexterous manipulation.

While I’m usually busy with my madness about robots, I’ve managed to pick up a few places where my heart is not at work. You can find me lost near trees and vast forests, which I believe is my natural habitat, writing (yes, I do, but I don’t show it), and cooking amazing food.

I’m always open to chat with people — drop me an email at offjangir@gmail.com with the subject “Hey Yash” :)

Scroll

Amazing People I Have Worked With :)

Selected Publications

Submitted

MotionForesight: Re-purposing Video Models for Future 3D Scene-Flow Prediction

Homanga Bharadhwaj*, Yash Jangir*

Submitted

SALT: semantically aligned action tokenization for VLAs
Preprint

Lost in Reconstruction: Aligning Action Representations with Language in Vision-Language-Action Models

Li Wenjie*, Yash Jangir, Ignacy Stepka, Yash Agarwal, Marion Kipsang, Yonatan Bisk

arXiv preprint, 2026

RobotArena project visualization
ICLR 2026 Accepted

RobotArena ∞: Unlimited Robot Benchmarking via Real-to-Sim Translation

Yash Jangir, Yidi Zhang*, Kashu Yamazaki*, Chenyu Zhang*, Kuan-Hsun Tu, Tsung-Wei Ke, Lei Ke, Yonatan Bisk, Katerina Fragkiadaki

International Conference on Learning Representations (ICLR), 2026

VideoManip: Dexterous manipulation from RGB human videos
Submitted

Dexterous Manipulation Policies from RGB Human Videos via 3D Hand-Object Trajectory Reconstruction

Hongyi Chen, Tony Dong, Tiancheng Wu, Liquan Wang, Yash Jangir, Yaru Niu, Yufei Ye, Homanga Bharadhwaj, Zackory Erickson, Jeffrey Ichnowski

Submitted

Visual servo control for aerial manipulation
IROS 2023

Image-based visual servo control for aerial manipulation using a fully‑actuated UAV

Guanqi He*, Yash Jangir*, Junyi Geng, Mohammadreza Mousaei, Dongwei Bai, Sebastian Scherer

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023

Education & Experience

Aug, 2026 – Present

Ph.D. in Computer Science

Johns Hopkins University

Advised by Prof. Homanga Bharadhwaj, working on robot learning, world models, and embodied AI.

May, 2026 – Present

Research Intern

Toyota Research Institute — Large Behavior Model Team

Research on large behavior models for generalist robot manipulation.

Aug, 2024 – May, 2026

M.S. Robotics

Carnegie Mellon University, Robotics Institute

Worked with Prof. Katerina Fragkiadaki and Prof. Yonatan Bisk on generalist robot policies, VLA models, and real-to-sim alignment.

May, 2022 – Jan, 2023

Robotics Institute Summer Scholar (RISS)

Carnegie Mellon University

Worked with Prof. Sebastian Scherer and Prof. Junyi Geng on aerial manipulation and visual control systems.

Aug, 2019 – May, 2023

B.E. Electronics and Instrumentation

BITS Pilani, Goa Campus

Undergraduate studies with focus on robotics, control systems, and computer vision.

Awards & Talks

RISS

RISS Scholar

Carnegie Mellon University

Selected as one of 50 students worldwide for the Robotics Institute Summer Scholars Program

IEEE

Best Paper Award

IEEE IRIA 2021

International Symposium on Intelligent Robotics and Industrial Automation

IEEE

Best Paper Finalist

IEEE ICMA 2022

International Conference on Mechatronics and Automation

Selected Projects

ImagineAct project visualization
Project

ImagineAct: Closed-Loop Vision-Language-Action with World Modeling

Yash Jangir*, Karan Mirakhor*, Tanya Choudhary*

Project (Sept 2025 - Present)

SpatialPi project visualization
Project

SpatialPi: Forcing Spatial Priors in Vision-Language-Action Models

Yidi Zhang*, Yash Jangir*

Project (Sept 2025 - Present)

ReMov3r project visualization
Project

ReMov3r: Real-Time Monocular Video to 3D Reconstruction

Yash Jangir*, Karan Mirakhor*, Tanya Choudhary*

Project (Sept 2024 - Dec 2024)

3D Scene Inpainting visualization
Project

3D Scene Inpainting with Text-to-Image Models and ControlNet

Yash Jangir

Project (Oct 2025 - Nov 2025)

Other Publications

Autonomous landing in MAVs
IRIA 2021 Best Paper

A generalized Kalman filter augmented deep‑learning approach for autonomous landing in MAVs

Pranay Mathur, Yash Jangir, Neena Goveas

Int. Symposium of Asian Control Association on Intelligent Robotics and Industrial Automation (IRIA), 2021

Force control on spherical manipulator
ICMA 2022 Best Paper Finalist

Implementation and testing of force control on a spherical articulated manipulator

Prathamesh Saraf, Yash Jangir, R. N. Ponnalagu

IEEE International Conference on Mechatronics and Automation (ICMA), 2022

3D pure pursuit guidance
ASCC 2022

3D pure pursuit guidance of drones for autonomous precision landing

Kaushal Kishore, Sagar Dalai, Yash Jangir, Samarth Singh, M Rohan, D Shashank, Sree S S Katta, S K Saha

13th Asian Control Conference (ASCC), 2022

Writing

Service

Reviewer

International Conference on Learning Representations (ICLR)

Reviewer

Conference on Neural Information Processing Systems (NeurIPS)

Reviewer

Conference on Robot Learning (CoRL)

Reviewer

IEEE Robotics and Automation Letters (RA-L)

Reviewer

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

"I'll show you that even someone without talent can succeed through effort."
— Rock Lee