Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Add aleatoric uncertainty to DenseNet (ablation) | DenseNet · CamVid | IoU: 67.1 -> 67.4 | Helped |
| Add epistemic uncertainty to DenseNet (ablation) | DenseNet · CamVid | IoU: 67.1 -> 67.2 | Helped |
| Add both aleatoric and epistemic uncertainty (ablation) | DenseNet · CamVid | IoU: 67.1 -> 67.5 | Helped |
| Add aleatoric uncertainty to DeepLabLargeFOV (ablation) | DeepLabLargeFOV · NYUv2 40-class | IoU: 36.5 -> 37.1 | Helped |
| Add epistemic uncertainty to DeepLabLargeFOV (ablation) | DeepLabLargeFOV · NYUv2 40-class | IoU: 36.5 -> 36.7 | Helped |
| Add both aleatoric and epistemic uncertainty (ablation) | DeepLabLargeFOV · NYUv2 40-class | IoU: 36.5 -> 37.3 | Helped |
| Add aleatoric uncertainty to DenseNet (ablation) | DenseNet · Make3D | rel: 0.167 -> 0.149 | Helped |
| Add aleatoric uncertainty to DenseNet (ablation) | DenseNet · Make3D | rms: 3.92 -> 3.93 | No clear effect |
| Add epistemic uncertainty to DenseNet (ablation) | DenseNet · Make3D | rel: 0.167 -> 0.162 | Helped |
| Add both aleatoric and epistemic uncertainty (ablation) | DenseNet · Make3D | rms: 3.92 -> 4.08 | Hurt |
| Add aleatoric uncertainty to DenseNet (ablation) | DenseNet · NYUv2 Depth | rel: 0.117 -> 0.112 | Helped |
| Add epistemic uncertainty to DenseNet (ablation) | DenseNet · NYUv2 Depth | rel: 0.117 -> 0.114 | Helped |
Closest papers by meaning
- Is Epistemic Uncertainty Faithfully Represented by Evidential Deep Learning Methods? (ICML 2024 · similarity 0.88)
- One Step Closer to Unbiased Aleatoric Uncertainty Estimation (AAAI 2024 · similarity 0.87)
- Benchmarking Uncertainty Disentanglement: Specialized Uncertainties for Specialized Tasks (NeurIPS 2024 · similarity 0.86)
- FUSE: Quantifying Uncertainty in Vision-Language Models by Bayesian Fusing Epistemic and Aleatoric Uncertainty (ICML 2026 · similarity 0.86)
- Are Uncertainty Quantification Capabilities of Evidential Deep Learning a Mirage? (NeurIPS 2024 · similarity 0.86)
- Epistemic Uncertainty for Generated Image Detection (NeurIPS 2025 · similarity 0.86)
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.