Research map / Label-efficient and robust learning / Recognition and long-tail
Generalized One-Class Learning Using Pairs of Complementary Classifiers
Anoop Cherian, Jue Wang
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 |
|---|---|---|---|
| add L2 unit normalization of input features (ablation) | GODS · JHMDB | F1 (%): 75.2 -> 77.7 | Improved |
| add L2 unit normalization of input features (ablation) | GODS · UCF-Crime | AUC (%): 68.9 -> 70.0 | Improved |
| add L2 unit normalization of input features (ablation) | GODS · Dash-Cam-Pose (TCN) | F1bar (%): 56.6 -> 58.4 | Improved |
| add L2 unit normalization of input features (ablation) | GODS_E (Euclidean) · UCF-Crime | AUC (%): 64.6 -> 66.8 | Improved |
| add soft-orthogonality penalty C with L2 normalization (ablation) | GODS · JHMDB | F1 (%): 77.7 -> 77.5 | No clear effect |
| add soft-orthogonality penalty C with L2 normalization (ablation) | GODS · UCF-Crime | AUC (%): 70.0 -> 70.3 | No clear effect |
| use Stiefel manifold instead of non-compact Stiefel (ablation) | GODS · UCF-Crime | AUC (%): 68.6 -> 70.0 | Improved |
| use Stiefel manifold instead of Euclidean manifold (ablation) | GODS · UCF-Crime | AUC (%): 66.8 -> 70.0 | Improved |
| use Stiefel manifold instead of oblique manifold (ablation) | GODS · UCF-Crime | AUC (%): 66.7 -> 70.0 | Improved |
| use Stiefel manifold instead of Euclidean manifold (ablation) | GODS · Dash-Cam-Pose (TCN) | F1bar (%): 55.8 -> 58.4 | Improved |
| use multiple orthonormal hyperplanes (GODS) instead of a single hyperplane pair (BODS) | GODS · JHMDB | F1: 0.725 -> 0.777 | Improved |
| use kernelized KODS instead of GODS | KODS · Dash-Cam-Pose (TCN) | F1bar: 0.584 -> 0.664 | Improved |
Closest papers by meaning
- GODS: Generalized One-Class Discriminative Subspaces for Anomaly Detection (ICCV 2019 · similarity 0.91)
- Learning and Evaluating Representations for Deep One-Class Classification (ICLR 2021 · similarity 0.87)
- Subspace Regularizers for Few-Shot Class Incremental Learning (ICLR 2022 · similarity 0.85)
- CLEAR: Cumulative LEARning for One-Shot One-Class Image Recognition (CVPR 2018 · similarity 0.85)
- Class Distribution Shifts in Zero-Shot Learning: Learning Robust Representations (NeurIPS 2024 · similarity 0.84)
- Polynomial-Time Robust Multiclass Linear Classification under Gaussian Marginals (NeurIPS 2026 · similarity 0.84)
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.
