AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Modeling Multimodal Aleatoric Uncertainty in Segmentation with Mixture of Stochastic Experts

Zhitong Gao, Yucong Chen, Chuyu Zhang, Xuming He

ICLR 2023 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Epistemic versus aleatoric decomposition · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use weighted-sample compact representation instead of standard samplingMoSE · LIDC-IDRI, all labelsGED (16): 0.218 $\pm$ 0.003 -> 0.195 $\pm$ 0.005Improved
Use weighted-sample compact representation instead of standard samplingMoSE · LIDC-IDRI, all labelsM-IoU (16): 0.624 $\pm$ 0.004 -> 0.635 $\pm$ 0.003Improved
Use weighted-sample compact representation instead of standard samplingMoSE · Multimodal Cityscapes, additional input, 35 samplesM-IoU: 0.690 $\pm$ 0.002 -> 0.761 $\pm$ 0.002Improved
Replace OT-based loss with pairwise IoU loss (ablation)MoSE · LIDC-IDRI, all labelsGED: 0.218 $\pm$ 0.003 -> 0.533 $\pm$ 0.001Hurt
Replace OT-based loss with pairwise IoU loss (ablation)MoSE · LIDC-IDRI, all labelsM-IoU: 0.624 $\pm$ 0.004 -> 0.533 $\pm$ 0.001Hurt
Replace OT-based loss with pairwise IoU loss (ablation)MoSE · LIDC-IDRI, all labelsECE: 0.064 $\pm$ 0.015 -> 0.277 $\pm$ 0.017Hurt
Use uniform instead of learnable expert weights (ablation)MoSE · LIDC-IDRI, all labelsGED: 0.218 $\pm$ 0.003 -> 0.282 $\pm$ 0.002Hurt
Use uniform instead of learnable expert weights (ablation)MoSE · LIDC-IDRI, all labelsM-IoU: 0.624 $\pm$ 0.004 -> 0.545 $\pm$ 0.007Hurt
Use uniform instead of learnable expert weights (ablation)MoSE · LIDC-IDRI, all labelsECE: 0.064 $\pm$ 0.015 -> 0.215 $\pm$ 0.006Hurt
Replace stochastic experts with deterministic experts (ablation)MoSE · LIDC-IDRI, all labelsGED: 0.218 $\pm$ 0.003 -> 0.246 $\pm$ 0.006Hurt
Replace stochastic experts with deterministic experts (ablation)MoSE · LIDC-IDRI, all labelsM-IoU: 0.624 $\pm$ 0.004 -> 0.591 $\pm$ 0.001Hurt
Replace stochastic experts with deterministic experts (ablation)MoSE · LIDC-IDRI, all labelsECE: 0.064 $\pm$ 0.015 -> 0.142 $\pm$ 0.003Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy