Nicolas Caytuiro's Doctoral Qualifying Exam

Topic: Symmetrization of 3D Generative Models

Abstract: We propose a novel data-centric approach to promote symmetry in 3D generative models by modifying the training data rather than the model architecture. Our method begins with an analysis of reflectional symmetry in both real-world 3D shapes and samples generated by state-of-the-art models. We hypothesize that training a generative model exclusively on half-objects, obtained by splitting shapes along the x=0 plane, enables the model to learn a rich distribution of partial geometries which, when reflected, yield complete shapes that are both visually plausible and geometrically symmetric. To test this, we construct a new dataset of half-objects from three ShapeNet classes (Airplane, Car, and Chair) and train a generative model using only the positive-$x$ side of each shape. Experiments demonstrate that the generated shapes are symmetrical and consistent, compared with generated objects from the original model and the original dataset objects.

Details

• Advisor: Ivan Sipiran
• Room: Picarte Auditorium
• Date: Monday, August 4th at 11:00 AM