小样本学习论文趋势 2023–2026

主要结论

• 小样本学习论文从 2023 年的 251 篇降到 2026 年的 224 篇，热度在回落。
• CVPR 是发小样本学习论文最多的会议，2023–2026 年共 303 篇，占四个会议总量的 34%。
• 交叉方向里涨得最快的是扩散模型：同时属于两个主题的论文从 2023 年的 4 篇增至 2026 年的 24 篇。
• 在「元学习与迁移」的 5 个主题里，它的 2026 年论文数排第 4；同组增长最快的基础模型适配是 3.8 倍，小样本学习是 0.9 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026BAPM: Boundary-Aware Prompt Mining for Training-Free Few-Shot Medical Image Segmentation
• ICML 2026Bayes-inspired Integration of Pretrained Priors and Few-Shot Evidence for Few-Shot Classification
• ICLR 2026Adaptive Concept Discovery for Interpretable Few-Shot Text Classification
• CVPR 2026A Closer Look at Cross-Domain Few-Shot Object Detection Fine-Tuning Matters
• NeurIPS 2025A Few Moments Please: Scalable Graphon Learning via Moment Matching
• ICML 2025FedBEns: One-Shot Federated Learning based on Bayesian Ensemble
• ICLR 2025Dist Loss: Enhancing Regression in Few-Shot Region through Distribution Distance Constraint
• CVPR 20253D Prior Is All You Need: Cross-Task Few-shot 2D Gaze Estimation
• NeurIPS 2024Few-shot Algorithms for Consistent Neural Decoding (FALCON) Benchmark
• ICML 2024Accelerating Convergence in Bayesian Few-Shot Classification
• ICLR 2024Few-shot Hybrid Domain Adaptation of Image Generator
• CVPR 2024Deformable One-shot Face Stylization via DINO Semantic Guidance
• NeurIPS 2023Alignment with human representations supports robust few-shot learning
• ICML 2023A Category-theoretical Meta-analysis of Definitions of Disentanglement
• ICLR 2023Bidirectional Language Models Are Also Few-shot Learners
• CVPR 2023A Meta-Learning Approach to Predicting Performance and Data Requirements
• NeurIPS 2026Beyond Normal References: Discriminative Few-Shot Anomaly Detection
• ICML 2026BiCrossNet with Decoupled Dual Generators: A Parameter‑Efficient and Generalizable Few‑Shot Custom Gesture Recognition Framework
• ICLR 2026Dual Distillation for Few-Shot Anomaly Detection
• CVPR 2026Beyond Patches Global-aware Autoregressive Model for Multimodal Few-Shot Font Generation

全部小样本学习论文

共 203 篇，第 2 / 5 页

51. CVPR 2025The Devil is in Low-Level Features for Cross-Domain Few-Shot Segmentation
52. CVPR 2025Token Cropr: Faster ViTs for Quite a Few Tasks
53. CVPR 2025Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning
54. CVPR 2025Twinner: Shining Light on Digital Twins in a Few Snaps
55. CVPR 2025UNEM: UNrolled Generalized EM for Transductive Few-Shot Learning
56. CVPR 2025UniGoal: Towards Universal Zero-shot Goal-oriented Navigation
57. CVPR 2025UniVAD: A Training-free Unified Model for Few-shot Visual Anomaly Detection
58. CVPR 2025Unleashing In-context Learning of Autoregressive Models for Few-shot Image Manipulation
59. CVPR 2025Visual and Semantic Prompt Collaboration for Generalized Zero-Shot Learning
60. CVPR 2025Z-Magic: Zero-shot Multiple Attributes Guided Image Creator
61. CVPR 2025Zero-shot 3D Question Answering via Voxel-based Dynamic Token Compression
62. CVPR 2025Zero-Shot Image Restoration Using Few-Step Guidance of Consistency Models (and Beyond)
63. CVPR 2025Zero-Shot Monocular Scene Flow Estimation in the Wild
64. CVPR 2025ZeroGrasp: Zero-Shot Shape Reconstruction Enabled Robotic Grasping
65. ICLR 2025Aligning Language Models with Demonstrated Feedback
66. ICLR 2025Black Sheep in the Herd: Playing with Spuriously Correlated Attributes for Vision-Language Recognition
67. ICLR 2025CapeX: Category-Agnostic Pose Estimation from Textual Point Explanation
68. ICLR 2025Dist Loss: Enhancing Regression in Few-Shot Region through Distribution Distance Constraint
69. ICLR 2025Divergence-enhanced Knowledge-guided Context Optimization for Visual-Language Prompt Tuning
70. ICLR 2025Efficient and Context-Aware Label Propagation for Zero-/Few-Shot Training-Free Adaptation of Vision-Language Model
71. ICLR 2025Emergence of meta-stable clustering in mean-field transformer models
72. ICLR 2025Federated Few-Shot Class-Incremental Learning
73. ICLR 2025FedTMOS: Efficient One-Shot Federated Learning with Tsetlin Machine
74. ICLR 2025From Few to Many: Self-Improving Many-Shot Reasoners Through Iterative Optimization and Generation
75. ICLR 2025Intermediate Layer Classifiers for OOD generalization
76. ICLR 2025Large (Vision) Language Models are Unsupervised In-Context Learners
77. ICLR 2025Learning to Adapt Frozen CLIP for Few-Shot Test-Time Domain Adaptation
78. ICLR 2025Local-Prompt: Extensible Local Prompts for Few-Shot Out-of-Distribution Detection
79. ICLR 2025LOIRE: LifelOng learning on Incremental data via pre-trained language model gRowth Efficiently
80. ICLR 2025Making Text Embedders Few-Shot Learners
81. ICLR 2025Manifold Induced Biases for Zero-shot and Few-shot Detection of Generated Images
82. ICLR 2025Meta-Dynamical State Space Models for Integrative Neural Data Analysis
83. ICLR 2025Metalic: Meta-Learning In-Context with Protein Language Models
84. ICLR 2025Multi-Perspective Data Augmentation for Few-shot Object Detection
85. ICLR 2025Multimodality Helps Few-shot 3D Point Cloud Semantic Segmentation
86. ICLR 2025One-for-All Few-Shot Anomaly Detection via Instance-Induced Prompt Learning
87. ICLR 2025PEAR: Primitive Enabled Adaptive Relabeling for Boosting Hierarchical Reinforcement Learning
88. ICLR 2025Predicate Hierarchies Improve Few-Shot State Classification
89. ICLR 2025Preserving Deep Representations in One-Shot Pruning: A Hessian-Free Second-Order Optimization Framework
90. ICLR 2025Protein Language Model Fitness is a Matter of Preference
91. ICLR 2025Prototype antithesis for biological few-shot class-incremental learning
92. ICLR 2025Safety Alignment Should be Made More Than Just a Few Tokens Deep
93. ICLR 2025STRAP: Robot Sub-Trajectory Retrieval for Augmented Policy Learning
94. ICLR 2025SynQ: Accurate Zero-shot Quantization by Synthesis-aware Fine-tuning
95. ICLR 2025UniMatch: Universal Matching from Atom to Task for Few-Shot Drug Discovery
96. ICLR 2025Unsupervised Zero-Shot Reinforcement Learning via Dual-Value Forward-Backward Representation
97. ICLR 2025Weighted Multi-Prompt Learning with Description-free Large Language Model Distillation
98. ICLR 2025Why In-Context Learning Models are Good Few-Shot Learners?
99. ICLR 2025Zero-Shot Natural Language Explanations
100. ICLR 2025ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning

常见问题

2023–2026 年四大会议一共有多少篇小样本学习论文？

共 898 篇，按年份分别是 2023 年 251 篇、2024 年 220 篇、2025 年 203 篇、2026 年 224 篇。

NeurIPS 2026 有多少篇小样本学习论文？

NeurIPS 2026 有 51 篇小样本学习论文，上一届是 68 篇。

ICLR 2026 有多少篇小样本学习论文？

ICLR 2026 有 38 篇小样本学习论文，上一届是 36 篇。

ICML 2026 有多少篇小样本学习论文？

ICML 2026 有 56 篇小样本学习论文，上一届是 35 篇。

CVPR 2026 有多少篇小样本学习论文？

CVPR 2026 有 79 篇小样本学习论文，上一届是 64 篇。

哪个会议的小样本学习论文最多？

CVPR，2023–2026 年共 303 篇。