Manan Shah

mt-shah@hotmail.com / [CV]

I am a master's student in computer vision (MSCV) at Carnegie Mellon University, and currently a Machine Learning Engineer Intern at Adobe (Firefly Foundry). My recent work has been in video diffusion models and 3D/4D reconstruction.
Previously, I was a research assistant at the Vision and AI Lab, IISc, under the mentorship of Prof. Venkatesh Babu, working on diffusion models and 3D Gaussian splatting.

I am graduating in December 2026 and actively looking for full-time Machine Learning Engineer or Research Engineer roles in perception.

I also maintain a blog where I jot down learnings, experiences, and ideas—writing helps me clarify concepts and share insights with others.
Feel free to reach out if you'd like to discuss interesting problems or opportunities!

News

[Jul 26] Our work, Lift4D, got accepted to SIGGRAPH Asia 2026.
[May 26] Joined Adobe as a Machine Learning Engineer Intern on the Firefly Foundry team.
[Aug 25] Started my master's in Computer Vision (MSCV) at CMU RI.
[Feb 25] Our work, MirrorVerse, on enhanced generation of mirror reflections got accepted to CVPR 2025.
[Nov 24] Our work, Reflecting Reality, on generating consistent mirror reflections got accepted to 3DV 2025.
[Mar 24] Joined Vision and AI Lab, IISc as a Research Assistant.
[Jul 23] Joined Cashfree Payments as a Software Development Engineer.
[Jul 23] Won the NCVPRIPG'23 challenge on writer verification.
[Jun 23] Graduated with a B.Tech in CSE from IIT Mandi.
[Jun 22] Interned with the Microsoft Search Technology Center, India (STCI) team.

Publications

2026
	Lift4D: Harmonizing Single-View 3D Estimation for 4D Reconstruction In-the-Wild Yehonathan Litman, Xiaoxuan Ma, Manan Shah, Nicolás Ugrinovic, Kris Kitani, Fernando De la Torre, Shubham Tulsiani* SIGGRAPH Asia 2026 [project page] [paper] [code]
2025
	MirrorVerse: Pushing Diffusion Models to Realistically Reflect the World Ankit Dhiman, Manan Shah, R Venkatesh Babu CVPR 2025 [project page] [paper] [code]
	Reflecting Reality: Enabling Diffusion Models to Produce Faithful Mirror Reflections Ankit Dhiman, Manan Shah, Rishubh Parihar, Yash Bhalgat, Lokesh R Boregowda, R Venkatesh Babu 3DV 2025 [project page] [paper] [code]