Oisin Mac Aodha

12

Papers

35

Total Citations

Papers (12)

MVSAnywhere: Zero-Shot Multi-View Stereo

DepthCues: Evaluating Monocular Depth Perception in Large Vision Models

AirPlanes: Accurate Plane Estimation via 3D-Consistent Embeddings

Representational Similarity via Interpretable Visual Concepts

The Automated LLM Speedrunning Benchmark: Reproducing NanoGPT Improvements

Improving Semantic Correspondence with Viewpoint-Guided Spherical Maps

SAOR: Single-View Articulated Object Reconstruction

From Coarse to Fine-Grained Open-Set Recognition

CrossSDF: 3D Reconstruction of Thin Structures From Cross-Sections

WildSAT: Learning Satellite Image Representations from Wildlife Observations

Representational Difference Explanations

CleverBirds: A Multiple-Choice Benchmark for Fine-grained Human Knowledge Tracing