Scholar
Matheus Gadelha
Google Scholar ID: VhqmvXsAAAAJ
Adobe Research
Citations & Impact
Citations
1,528
H-index
14
i10-index
17
Publications
20
Co-authors
24
list available
Publications
Resume (English only)
Academic Achievements
- Published several papers, including SIGMA-Gen: Structure and Identity Guided Multi-subject Assembly for Image Generation, Frame In-N-Out: Unbounded Controllable Image-to-Video Generation, Reusing Computation in Text-to-Image Diffusion for Efficient Generation of Image Sets, DMesh++: An Efficient Differentiable Mesh for Complex Shapes, 3D-Fixup: Advancing Photo Editing with 3D Priors, PreciseCam: Precise Camera Control for Text-to-Image Generation, Motion Modes: What Could Happen Next?, Instant3dit: Multiview Inpainting for Fast Editing of 3D Objects, Text-guided Controllable Mesh Refinement for Interactive 3D Modeling, DMesh: A Differentiable Representation for General Meshes, GEM3D: Generative Medial Abstractions for 3D Shape Synthesis, Learning Continuous 3D Words for Text-to-Image Generation, etc.
Research Experience
- Worked on multiple projects at Adobe Research, including Adobe Illustrator Mockup (applying 2D designs to real photographs in a 3D-aware manner), Adobe's Project Neo (creating 3D shapes easily and rendering them using text-to-image generative models), and Adobe Substance Viewer (using image generative models to 'render' a 3D scene according to a text prompt and style presets).
Education
- PhD from the University of Massachusetts - Amherst, supervised by Prof. Rui Wang and Prof. Subhransu Maji.
Background
- Currently a Research Scientist at Adobe Research. Interested in Computer Graphics, Vision, and their intersections with Machine Learning. Focus on models and representations of tridimensional data, especially incorporating 3D capabilities into large generative models for better control and understanding.
Miscellany
- Open to academic collaborations; feel free to reach out via email to discuss potential research problems.
