Research map / Label-efficient and robust learning / OOD and anomaly detection
GradOrth: A Simple yet Efficient Out-of-Distribution Detection with Orthogonal Projection of Gradients
Sima Behpour, Thang Doan, Xin Li, Wenbin He, Liang Gou, Liu Ren
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use last layer instead of all layers (ablation) | GradOrth · ImageNet ResNet-50 | avg FPR95: 20.35 -> 18.57 | Improved |
| use last layer instead of all layers (ablation) | GradOrth · ImageNet MobileNetV2 | avg FPR95: 29.17 -> 27.65 | Improved |
| use last layer instead of all layers (ablation) | GradOrth · CIFAR-10 | avg FPR95: 13.16 -> 12.34 | Improved |
| use last layer instead of all layers (ablation) | GradOrth · CIFAR-100 | avg FPR95: 34.84 -> 33.35 | Improved |
| use last layer instead of all layers (ablation) | GradOrth · ImageNet MobileNetV2 iNaturalist | FPR95: 26.14 -> 26.81 | Hurt |
| increase samples per class for subspace from 5 to 10 (ablation) | GradOrth · CIFAR-10 | avg FPR95: 12.34 -> 11.86 | No clear effect |
| increase samples per class for subspace from 5 to 20 (ablation) | GradOrth · CIFAR-10 | avg FPR95: 12.34 -> 11.90 | No clear effect |
| increase samples per class for subspace from 5 to 40 (ablation) | GradOrth · CIFAR-100 | avg FPR95: 33.35 -> 32.92 | No clear effect |
| increase samples per class for subspace from 5 to 10 (ablation) | GradOrth · CIFAR-100 | avg FPR95: 33.35 -> 33.34 | No clear effect |
| use L2 norm instead of other Lp norms (L1, L3, L4, fraction) (ablation) | GradOrth · ImageNet | FPR95 and AUROC | Improved |
| use L-infinity norm instead of L2 (ablation) | GradOrth · ImageNet | FPR95 and AUROC | Hurt |
| fractional norm (p=0.3) instead of L2 (ablation) | GradOrth · ImageNet | FPR95 and AUROC | No clear effect |
Later papers that compared with this method
2 later accepted papers report GradOrth as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Fisher-Rao Sensitivity for Out-of-Distribution Detection in Deep Neural Networks (ICLR 2026 · better on 5 of 8 shared results)
- Kernel PCA for Out-of-Distribution Detection: Non-Linear Kernel Selection and Approximation (TPAMI 2026 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Double Descent Meets Out-of-Distribution Detection: Theoretical Insights and Empirical Analysis on the Role of Model Complexity (NeurIPS 2025 · similarity 0.89)
- Mining In-distribution Attributes in Outliers for Out-of-distribution Detection (AAAI 2025 · similarity 0.89)
- Learning to Shape In-distribution Feature Space for Out-of-distribution Detection (NeurIPS 2024 · similarity 0.89)
- Test-Time Linear Out-of-Distribution Detection (CVPR 2024 · similarity 0.89)
- Spurious-Aware Prototype Refinement for Reliable Out-of-Distribution Detection (NeurIPS 2025 · similarity 0.89)
- Discriminability-Driven Channel Selection for Out-of-Distribution Detection (CVPR 2024 · similarity 0.89)
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.
