CVPR 2026 accepted papers on Label-efficient and robust learning
283 of the 4,042 papers accepted at CVPR 2026 (7.0%) are on Label-efficient and robust learning, compared with 7.7% at CVPR 2025. They fall into 9 topics.
See CVPR 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Recognition and long-tail · 95 papers
- ActiveAD: Planning-Oriented Active Learning for End-to-End Autonomous Driving (2 citations)
- A2GC: Asymmetric Aggregation with Geometric Constraints for Locally Aggregated Descriptors
- Active Inference for Micro-Gesture Recognition: EFE-Guided Temporal Sampling and Adaptive Learning
- BarbieGait: An Identity-Consistent Synthetic Human Dataset with Versatile Cloth-Changing for Gait Recognition
- BD-Merging: Bias-Aware Dynamic Model Merging with Evidence-Guided Contrastive Learning
- Beyond Top Activations: Efficient and Reliable Crowdsourced Evaluation of Automated Interpretability
- Bias In, Bias Out? Finding Unbiased Subnetworks in Vanilla Models
- BIT: Matching-based Bi-directional Interaction Transformation Network for Visible-Infrared Person Re-Identification
- Cleaning the Pool: Progressive Filtering of Unlabeled Pools in Deep Active Learning
- Confusion-Aware Spectral Regularizer for Long-Tailed Recognition
- CUE: Concept-Aware Multi-Label Expansion to Mitigate Concept Confusion in Long-Tailed Learning
- Decision Boundary-aware Generation for Long-tailed Learning
and 83 more in the interactive map.
Domain adaptation and generalization · 41 papers
- $\oslash$ Source Models Leak What They Shouldn't $\nrightarrow$: Unlearning Zero-Shot Transfer in Domain Adaptation Through Adversarial Optimization
- Addressing Exacerbated Attention Sink for Source-Free Cross-Domain Few-Shot Learning
- Bridging Domains through Subspace-Aware Model Merging
- CHIPS: Efficient CLIP Adaptation via Curvature-aware Hybrid Influence-based Data Selection
- CoFiDA-M: Concept-Aware Feature Modulation for Cross-Domain Adaptation with Image-Only Inference
- Cross-Domain Demo-to-Code via Neurosymbolic Counterfactual Reasoning
- Cross-Domain Few-Shot Segmentation via Multi-view Progressive Adaptation
- CrossEarth-Gate: Fisher-Guided Adaptive Tuning Engine for Efficient Adaptation of Cross-Domain Remote Sensing Semantic Segmentation
- Denoise and Align: Towards Source-Free UDA for Robust Panoramic Semantic Segmentation
- Heuristic Self-Paced Learning for Domain Adaptive Semantic Segmentation under Adverse Conditions
- Interpretable Cross-Domain Few-Shot Learning with Rectified Target-Domain Local Alignment
- Learnable Motion-Focused Tokenization for Effective and Efficient Video Unsupervised Domain Adaptation
and 29 more in the interactive map.
OOD and anomaly detection · 36 papers
- Bidirectional Multimodal Prompt Learning with Scale-Aware Training for Few-Shot Multi-Class Anomaly Detection (1 citation)
- Activation Matters: Test-time Activated Negative Labels for OOD Detection with Vision-Language Models
- Adaptive Confidence Regularization for Multimodal Failure Detection
- Enhancing Out-of-Distribution Detection with Extended Logit Normalization
- GS-CLIP: Zero-shot 3D Anomaly Detection by Geometry-Aware Prompt and Synergistic View Representation Learning
- Hierarchical Point-Patch Fusion with Adaptive Patch Codebook for 3D Shape Anomaly Detection
- Image-based Outlier Synthesis With Training Data
- Inside-Out: Measuring Generalization in Vision Transformers Through Inner Workings
- KLIP: Localized Distribution Shift Detection via KL-Divergence with Diffusion Priors in Inverse Problems
- Mind the Way You Select Negative Texts: Pursuing the Distance Consistency in OOD Detection with VLMs
- Mitigating Simplicity Bias in OOD Detection through Object Co-occurrence Analysis
- MMR-AD: A Large-Scale Multimodal Dataset for Benchmarking General Anomaly Detection with Multimodal Large Language Models
and 24 more in the interactive map.
Test-time adaptation · 26 papers
- AcTTA: Rethinking Test-Time Adaptation via Dynamic Activation
- Adapting Point Cloud Analysis via Multimodal Bayesian Distribution Learning
- Anatomical Domain Shifts: Test-time Heterogeneous Adaptation for 3D Human Pose Prediction
- Back to Source: Open-Set Continual Test-Time Adaptation via Domain Compensation
- Bridging Domain Expertise and Generalization for Performance Estimation
- CD-Buffer: Complementary Dual-Buffer Framework for Test-Time Adaptation in Adverse Weather Object Detection
- Curvature-Aware Zeroth-Order Optimization for Memory-Efficient Test-Time Adaptation
- Dance Across Shifts: Forward-Facilitation Continual Test-Time Adaptation through Dynamic Style Bridging
- Decoupling Stability and Plasticity for Multi-Modal Test-Time Adaptation
- Dual-level Adaptation for Multi-Object Tracking: Building Test-Time Calibration from Experience and Intuition
- FOZO: Forward-Only Zeroth-Order Prompt Optimization for Test-Time Adaptation
- Multi-modal Test-time Adaptation via Adaptive Probabilistic Gaussian Calibration
and 14 more in the interactive map.
Self-supervised and contrastive · 24 papers
- OlmoEarth: Stable Latent Image Modeling for Multimodal Earth Observation (2 citations)
- E-RayZer: Self-supervised 3D Reconstruction as Spatial Visual Pre-training (1 citation)
- Beyond Binary Contrast: Modeling Continuous Skeleton Action Spaces with Transitional Anchors
- Boosting Self-Supervised Tracking with Contextual Prompts and Noise Learning
- CanonCGT: Reference-Based Color Grading via Canonical Pivot Representation
- DreamSAC: Learning Hamiltonian World Models via Symmetry Exploration
- Franca: Nested Matryoshka Clustering for Scalable Visual Representation Learning
- Intervention-Aware Multiscale Representation Learning from Imaging Phenomics and Perturbation Transcriptomics
- InvCoSS: Inversion-driven Continual Self-supervised Learning in Medical Multi-modal Image Pre-training
- Learning Convex Decomposition via Feature Fields
- Learning Effective Sign Features without Text for Gloss-free Sign Language Translation
- Learning Generalizable 3D Medical Image Representations from Mask-Guided Self-Supervision
and 12 more in the interactive map.
Knowledge distillation · 19 papers
- AMap: Distilling Future Priors for Ahead-Aware Online HD Map Construction
- Brewing Stronger Features: Dual-Teacher Distillation for Multispectral Earth Observation
- Continual Distillation of Teachers from Different Domains
- Dual-branch Distilled Transformer for Efficient Asymmetric UAV Tracking
- Fixed Anchors Are Not Enough: Dynamic Retrieval and Persistent Homology for Dataset Distillation
- Foundry: Distilling 3D Foundation Models for the Edge
- Generalizable Knowledge Distillation from Vision Foundation Models for Semantic Segmentation
- GeoSANE: Learning Geospatial Representations from Models, Not Data
- HAD: Heterogeneity-Aware Distillation for Lifelong Heterogeneous Learning
- LIFT and PLACE: A Simple, Stable, and Effective Knowledge Distillation Framework for Lightweight Diffusion Models
- Momentum Memory for Knowledge Distillation in Computational Pathology
- Rethinking Dataset Distillation: Hard Truths about Soft Labels
and 7 more in the interactive map.
Zero-/few-shot and meta-learning · 19 papers
- Fast-FoundationStereo: Real-Time Zero-Shot Stereo Matching
- GraspGen-X: Cross-Embodiment 6-DOF Diffusion-based Grasping
- Incentivizing Generative Zero-Shot Learning via Outcome-Reward Reinforcement Learning with Visual Cues
- LiDAS: Lighting-driven Dynamic Active Sensing for Nighttime Perception
- Lite Any Stereo: Efficient Zero-Shot Stereo Matching
- ProjFlow: Projection Sampling with Flow Matching for Zero-Shot Exact Spatial Motion Control
- PromptStereo: Zero-Shot Stereo Matching via Structure and Motion Prompts
- What Makes Good Synthetic Training Data for Zero-Shot Stereo Matching?
- Data-Centric Meta-Learning for Robust Few-Shot Generalization
- DDSF: Robust Few-Shot Learning via Disentangled Subspaces with Determinantal Point Process
- From Few-way to Many-way: Rethinking Few-shot Fine-grained Image Classification
- Graph Attention Prototypical Network for Robust Few-Shot Classification
and 7 more in the interactive map.
Noisy labels · 12 papers
- ConeSep: Cone-based Robust Noise-Unlearning Compositional Network for Composed Image Retrieval (1 citation)
- Batch Loss Score for Dynamic Data Pruning
- Deconstructing the Failure of Ideal Noise Correction: A Three-Pillar Diagnosis
- Drainage: A Unifying Framework for Addressing Class Uncertainty
- Revisiting Sparsity Constraint Under High-Rank Property in Partial Multi-Label Learning
- See Through the Noise: Improving Domain Generalization in Gaze Estimation
- A Supervised Multi-task Framework for Joint cryo-ET Restoration Enabled by Generative Physical Simulation
- CLEX: Complementary Label Exchange Learning for Noisy Facial Expression Recognition
- Debiased Sample Selection for Learning with Noisy Labels
- Dynamic Label Noise Suppression with Optimal Teacher Pool for Facial Expression Recognition
- Revisiting Learning with Noisy Labels: Active Forgetting and Noise Suppression
- TANGO: Text-Anchored Guided Optimization for Robust Fine-tuning Vision-Language Models under Label Noise
Semi-supervised learning · 11 papers
- LLM-Guided Probabilistic Fusion for Label-Efficient Document Layout Analysis
- Real-World Point Tracking with Verifier-Guided Pseudo-Labeling
- SECOS: Semantic Capture for Rigorous Classification in Open-World Semi-Supervised Learning
- SemiGDA: Generative Dual-distribution Alignment for Semi-Supervised Medical Image Segmentation
- Towards Multimodal Domain Generalization with Few Labels
- Divide, Conquer, and Aggregate: Asymmetric Experts for Class-Imbalanced Semi-Supervised Medical Image Segmentation
- From Softmax to Dirichlet: Evidential Learning for Semi-supervised Semantic Segmentation
- GaussianMatch: Semi-Supervised Regression with Pseudo-Label Filtering via Multi-View Gaussian Consistency
- PAF: Perturbation-Aware Filtering for Open-Set Semi-Supervised Learning
- Semi-supervised Echocardiography Video Segmentation via Anchor Semantic Awareness and Continuous Pseudo-label Reforging
- Spatial-SAM: Spatially Consistent 3D Electron Microscopy Segmentation with SDF Memory and Semi-Supervised Learning