Deva Ramanan

RF-DETR: Neural Architecture Search for Real-Time Detection Transformers

Contact-guided Real2Sim from Monocular Video with Planar Scene Primitives

UniFlow: Zero-Shot LiDAR Scene Flow via Cross-Domain Generalization

Scaling Short-Term Memory of Visuomotor Policies for Long-Horizon Tasks

Robotic Steering: Mechanistic Finetuning for Vision-Language-Action Models

Activation Reward Models for Few-Shot Model Alignment

Benchmarking Stereo Geometry Estimation in the Wild

UFM: A Simple Path towards Unified Dense Correspondence with Flow

RaySt3R: Predicting Novel Depth Maps for Zero-Shot Object Completion

Neural Eulerian Scene Flow Fields

TAO-Amodal: A Benchmark for Tracking Any Object Amodally

Self-Correcting Decoding with Generative Feedback for Mitigating Hallucinations in Large Vision-Language Models

Reconstruct, Inpaint, Test-Time Finetune: Dynamic Novel-view Synthesis from Monocular Videos