About Me

I am currently a Staff Research Scientist at the Toyota Research Institute, where I lead the Geometric 3D Vision research direction, as well as oversee several efforts towards the productization of this research, and participate in university collaborations. This includes self-supervised learning, depth estimation, optical flow, scene flow, keypoints, multi-view geometry, camera calibration, and so forth. I received my PhD at the University of Sydney, under the supervision of Prof. Fabio Ramos, and did my PostDoc studies at the same institution. When I am not doing all these things, I enjoy spending time with my wife, playing the flute (mostly transverse and tin whistle), role playing games (I wrote a book about my D&D campaign, if you read Portuguese you can buy it here), and reading fantasy and sci-fi novels (huge Tolkien, Sanderson, Egan, and Baxter fan).


News

12/10/2024: Attending NeurIPS

I'll be attending NeurIPS in Vancouver this year! Some highlights of what I'll be doing:

* TRI will have a booth, so if you would like to learn more about what we're up to these days make sure to stop by!

* I'll be presenting Score Distillation via Reparametrized DDIM , done in collaboration with MIT as part of our University 2.0 program (Thu, Dec 12, 11:00, 3 East).

* I'll be presenting SE(3) Equivariant Ray Embeddings for Implicit Multi-View Depth Estimation , done in collaboration with UPenn as part of the TRI internship program (Fri, Dec 13, 16:30, 6 East).

11/05/2024: 3DV Accepted Papers

Happy to announce two papers accepted to 3DV 2025!

* GRIN: Zero-Shot Metric Depth with Pixel-Level Diffusion

* Incorporating Dense Metric Depth into Neural 3D Representations for View Synthesis and Relighting

09/29/2024: Papers at ECCV

Unfortunately I won't be able to make it to ECCV this year, but if you are around make sure to look up these papers!

* NeRF-MAE: Masked AutoEncoders for Self-Supervised 3D Representation Learning for Neural Radiance Fields

* Zero-Shot Multi-Object Scene Completion

09/25/2024: NeurIPS Accepted Papers

Happy to announce two papers accepted to NeurIPS 2024!

* SE(3) Equivariant Ray Embeddings for Implicit Multi-View Depth Estimation

* Score Distillation via Reparametrized DDIM

06/17/2024: Attending CVPR

I'll be attending CVPR in Seattle this year! Some highlights of what I'll be doing:

* I'm co-organizing the EquiVision Workshop. If you are interested in leveraging symmetry in structured data to improve generalization, data efficiency and robustness, please check it out!

* I'm giving a keynote talk at the MDEC Workshop, if you are interested in monocular depth estimation and how it has evolved over the years, please check it out! UPDATE: The slides for my talk can be found here.

* I'll be at the poster session for our latest paper: Towards Realistic Scene Generation with LiDAR Diffusion Models, so please stop by if you want to have a chat and learn more! It's on Thursday 5:00-6:30pm at Arch 4A-E, poster #12.