Papers
Single View Seafloor Recovery from Imaging Sonar via Differentiable Rendering
Sevan Brodjian, Michael Hobley, Pietro Perona
CVPR PBVS Workshop · Accepted · May 2026
Forward-looking sonar collapses the entire vertical structure of a 3D scene into a single flat, ambiguous image, which made inverting it an interesting problem. We built a fully differentiable renderer of the acoustic acquisition physics and let gradient descent do the rest, recovering seafloor and riverbed geometry from a single frame with no training data.
Paper · Project page · Code
Training-Free Temporal Abstraction for General Video Understanding
Etienne Cassanova, Sevan Brodjian, Pietro Perona
Under Review · May 2026
In this paper we demonstrate that pretrained foundation models provide a general video understanding backbone across moment retrieval, generic event boundary detection, and long-video frame selection for VLMs.
Kuramoto Orientation Diffusion Models
Yue Song, T. Anderson Keller, Sevan Brodjian et al.
NeurIPS 2025 · Accepted (Poster) · September 2025
Biological neural systems synchronize through Kuramoto oscillator dynamics. We copied that mechanism and built a diffusion model out of it, using phase synchronization as a structured prior for generating orientation-rich images like fingerprints and textures.