Hallison Paz

São Paulo, Brazil

I am an AI Graphics Researcher and PhD Candidate at IMPA 🇧🇷. My research interest lies in the intersection of computer graphics, computer vision, and machine learning. My PhD research focuses on neural implicit representations for media objects such as images, videos, and 3D models. In 2022, I was a Research Scientist Intern at Reality Labs Research (Meta) in Pittsburgh, PA 🇺🇸, where I contributed to the Codec Avatars project, working with inverse rendering, 3D modeling and generative models.

I am also the co-creator of the YouTube channel Programação Dinâmica, where I talk about programming, data science, artificial intelligence, and scientific outreach, along with my personal experiences and career as a researcher. The channel has garnered over 200K subscribers, was part of the YouTube Black Voices Fund class of 2022, and was recognized by YouTube as a Creator of Impact in 2024.

I have extensive experience in software development, teaching programming across all levels — from completely beginners to graduate students — and creating educational content in both text and video formats. I hold a Master’s degree in Mathematics with a focus on Computer Graphics from IMPA, where my thesis explored implicit surface reconstruction from depth image data. Additionally, I have a Bachelor’s degree in Computer Engineering from the Instituto Militar de Engenharia (Military Institute of Engineering).

If you want to shape new perspectives on the world, whether digitally or physically, please call me in.

news

Aug 18, 2024	My Youtube channel has reached 200.000 subscribers
Jul 27, 2024	I’m glad to be in Denver presenting my poster Spectral Periodic Networks for Neural Rendering at SIGGRAPH!
Jun 17, 2024	I am happy to be on CVPR 2024 presenting Neural Implicit Morphing of Face Images
Aug 23, 2023	I was invited to publish an extended version of the MR-Net research in the journal Computers & Graphics!
Oct 24, 2022	My first paper based on my PhD research has been published on SIBGRAPI - 35th Conference on Graphics, Patterns and Images.
Jul 05, 2022	I am very excited to be starting my internship at Meta, Reality Labs Research, in Pittsburgh.

selected publications

Spectral Periodic Networks for Neural Rendering

Hallison Paz, Tiago Novello, and Luiz Velho

In ACM SIGGRAPH 2024 Posters , Denver, CO, USA, 2024
Neural Implicit Morphing of Face Images

Guilherme Schardong, Tiago Novello, Hallison Paz, Iurii Medvedev, and 3 more authors

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , Jun 2024
MR-Net: Multiresolution sinusoidal neural networks

Hallison Paz, Daniel Perazzo, Tiago Novello, Guilherme Schardong, and 6 more authors

Computers & Graphics, Jun 2023

Abs

We present MR-Net, a general architecture for multiresolution sinusoidal neural networks, and a framework for imaging applications based on this architecture. We extend sinusoidal networks, and we build an infrastructure to train networks to represent signals in multiresolution. Our coordinate-based networks, namely L-Net, M-Net, and S-Net, are continuous both in space and in scale as they are composed of multiple stages that progressively add finer details. Currently, band-limited coordinate networks (BACON) are able to represent signals at multiscale by limiting their Fourier spectra. However, this approach introduces artifacts leading to an image with a ringing effect. We show that MR-Net can represent more faithfully what is expected of sequentially applying low-pass filters in a high-resolution image. Our experiments on the Kodak Dataset show that MR-Net can reach comparable Peak Signal-to-Noise Ratio (PSNR) to other architectures, on image reconstruction, while needing fewer additional parameters for multiresolution. Along with MR-Net, we detail our architecture’s mathematical foundations and general ideas, and show examples of applications to texture magnification, minification, and antialiasing. Lastly, we compare our three MR-Net subclasses.
Multiresolution Neural Networks for Imaging

Hallison Paz, Tiago Novello, Vinicius Silva, Guilherme Shardong, and 4 more authors

In Proceedings of SIBGRAPI , Jun 2022