Minghe Shen — accepted AI-conference papers

2026

MD2E: Modeling Depth-to-Edge Cues for Monocular Metric Depth Estimation

CVPR 2026

We study monocular metric depth estimation (MMDE) without camera intrinsics at training or inference. When focal length and scene depth vary together, depth changes are difficult to perceive in the image, yet the edge-frequency statistics exhibit systematic, scale-correlated shifts. Building on this

Cited by 0 Source ↗code ↗Scholar ↗

2026

Noisy but Valid: Robust Statistical Evaluation of LLMs with Imperfect Judges

Chen Feng, Minghe Shen, Ananth Balashankar, Carsten Gerner-Beuerle, Miguel R. D. Rodrigues

ICLR 2026poster

Reliable certification of Large Language Models (LLMs)—verifying that failure rates are below a safety threshold—is critical yet challenging. While "LLM-as-a-Judge" offers scalability, judge imperfections, noise, and bias can invalidate statistical guarantees. We introduce a "Noisy but Valid" hypoth…

Cited by 0 Source ↗Scholar ↗