Maria Beatriz Silva

I am a Presidential Honors Scholar studying Computer Science at NYU’s Courant Institute with a minor in Mathematics. As a Research Assistant at NYU’s Immersive Computing Lab, I work on problems at the intersection of computer graphics, human perception, and machine learning. My work is driven by the commitment to ensure technology is developed with the needs of its users in mind and by the desire to reshape who gets to participate in and benefit from technological innovation.

Conference Publications

GeneVA: A Dataset of Human Annotations for Generative Text to Video Artifacts

GeneVA Dataset Overview

Our Artifact Taxonomy

Kang, J., Silva, M. B., Sangkloy, P., Chen, K., Williams, N., & Sun, Q. (arXiv Preprint, 2025)

Accepted to Winter Conference on Applications of Computer Vision (WACV) 2026

📝 arXiv 🎥 Video Presentation

Abstract: The first large-scale dataset of human-annotated artifacts in AI-generated videos, containing 16,451 annotations across 16,356 videos with per-frame bounding boxes, descriptions, and quality ratings. This work addresses the critical need for human-in-the-loop feedback for generative video models.

My Contributions:

  • Designed and implemented the complete data acquisition pipeline, developing a custom Python system to extract videos from 2.7TB dataset under HPC storage constraints
  • Conducted video analysis and literature review that informed the artifact taxonomy development
  • Proposed key study design decision to show prompts to annotators, enabling measurement of both prompt-video alignment and artifact perception

PaleoScan: Low-Cost Easy-to-use High-Volume Fossil Scanning

PaleoScan Teaser

Silva, C., Piadyk, Y., Rulff, J., Panozzo, D., Silva, M. B., et al. (2024)

ACM CHI Conference on Human Factors in Computing Systems

📄 Paper 🎥 Video Preview 💻 Full Video Presentation

Abstract: PaleoScan presents a low-cost, accessible 3D scanning system designed for high-volume fossil digitization in resource-limited paleontological institutions, democratizing access to advanced scanning technology.

My Contributions:

  • Collaborated with paleontologists in Brazil to understand workflows and system requirements
  • Helped conceptualize the interface design for PaleoDP, the data processing and annotation pipeline
  • Directed and produced the video submission
  • Co-presented the paper at ACM CHI 2024

Presentations

ACM CHI 2024 — Honolulu, Hawaii
Co-presented PaleoScan: Low-Cost Easy-to-use High-Volume Fossil Scanning
May 2024