Benjamin Attal

PhD student @CMU_Robotics, doing research in computer vision, computational imaging. imasquare was taken.

[Eurographics Symposium on Rendering @EGSympRendering]

EGSR'26 Call for Papers is out!

Doing research on rendering? Physics-based, neural, stylized... Submit your work to EGSR!
Papers deadline: April 8th 2026

https://t.co/FBsuahA1eP

[Andrew Guo @andrewyguo]

📢📢📢Dark3R (#CVPR2026) recovers camera poses and point clouds—and enables novel view synthesis—from images captured in ultra-low light (image SNR < −4 dB).

Project Page: https://t.co/Hea6ACUPCs
Code/Dataset: https://t.co/PFnrOuYIfB
Paper: https://t.co/8CGZXbS6t3

[CMU Robotics Institute @CMU_Robotics]

Congrats to RI Ph.D. Benjamin Attal, RI assistant professor Matthew O'Toole & co-authors from @UofT & @VectorInst for winning the CVPR 2025 Best Student Paper award! 🙌

Check out the winning work: "Neural Inverse Rendering from Propagating Light" 💡🏆

https://t.co/QCwrJYG1En

[David Lindell @DaveLindell]

Honored that our work received the best student paper award at #CVPR2025! This was a really fun and exciting collaboration with @mpotoole led by amazing students @anagh_malik @imarhombus @AndrewEJXie!

Check out the work at https://t.co/lSPS4WJ2EH

[#CVPR2026 @CVPR]

#CVPR2025 paper awards

[Benjamin Attal @imarhombus]

A reminder to please join us tomorrow (June 12) afternoon at #CVPR2025 in room 106 C for the first workshop on Physics-inspired 3D vision and Imaging!

[Anagh Malik @anagh_malik]

📢📢📢 Come and submit to our workshop on Physics-inspired 3D Vision and Imaging at CVPR 2025!

Speakers 🗣️ include Ioannis Gkioulekas, Laura Waller, Berthy Feng, @SeungHwanBaek8 and @GordonWetzstein!

Thanks to coorganizers @imarhombus, @ceciliazhang77, @dorverbin and @jtompkin!

[Benjamin Attal @imarhombus]

Check out our work "Neural Inverse Rendering from Propagating Light", appearing at CVPR 2025 as an oral presentation. It was awesome collaborating @anagh_malik -- super bright, and a super hard worker who made everything easy :)

[Anagh Malik @anagh_malik]

📢📢📢 Neural Inverse Rendering from Propagating Light 💡

Our CVPR Oral introduces the first method for multiview neural inverse rendering from videos of propagating light, unlocking applications such as relighting light propagation videos, geometry estimation, or light component separation!

\w @imarhombus (co-first), @AndrewEJXie, @mpotoole, and @DaveLindell

[Benjamin Attal @imarhombus]

Excited to be part of the organizing team for the first workshop on Physics-inspired 3D Vision and Imaging! If you do research at the intersection of physical modeling & 3D/4D reconstruction please consider submitting your work :)

Website: https://t.co/DnbUu23kfl

[Anagh Malik @anagh_malik]

📢📢📢 Come and submit to our workshop on Physics-inspired 3D Vision and Imaging at CVPR 2025!

Speakers 🗣️ include Ioannis Gkioulekas, Laura Waller, Berthy Feng, @SeungHwanBaek8 and @GordonWetzstein!

Thanks to coorganizers @imarhombus, @ceciliazhang77, @dorverbin and @jtompkin!

[Dor Verbin @dorverbin]

We’ll be presenting NeRF-Casting at SIGGRAPH Asia next week! NeRF-Casting enables photorealistic rendering of scenes with highly reflective surfaces—something that was previously impossible with models like Zip-NeRF and 3DGS. (1/6)

[Christian Richardt @c_richardt]

Learn how to interpolate time-of-flight depth videos using Flowed Time of Flight Radiance Fields (F-TöRF) at this morning’s @eccvconf poster session.

Project:
https://t.co/j6klPyQ9y4

With Mikhail Okunev*, Marc Mapeke*, @imarhombus, @mpotoole, @jtompkin

[Benjamin Attal @imarhombus]

@jbhuang0604 Hope you get well soon, Jia-Bin!

[Benjamin Attal @imarhombus]

@JonClark55 Yep metalness is part of the BRDF parameterization (Disney-GGX) that we use!

[Benjamin Attal @imarhombus]

(1/N) Flash Cache: Reducing Bias in Radiance Cache Based Inverse Rendering

Website: https://t.co/ytSXIoeyKo

tl;dr our #ECCV2024 (oral ✨) paper presents a new system for inverse rendering that is more physically accurate, and therefore less biased, than existing approaches.

[Benjamin Attal @imarhombus]

@taiuti @_pratul_ @dorverbin @BenMildenhall @PeterHedman3 @jon_barron @mpotoole Yes, radiance caches are quite like light probes! But rather than providing incoming light at a discrete set of locations/probes, a radiance cache gives incoming light at any point in the scene

[Benjamin Attal @imarhombus]

This work was carried out while I was an SR at Google, with amazing manager @_pratul_ and colleagues @dorverbin @BenMildenhall @PeterHedman3 @jon_barron, as well as my awesome PhD advisor
@mpotoole

[Benjamin Attal @imarhombus]

(9/N) Our full system for inverse rendering combines the high-quality radiance cache, occlusion-aware importance sampling, and fast cache to produce low--cost and accurate rendering --- while still treating scene geometry as volumetric.