mlsanityData to 7 Oct 2026

PMCE: Probabilistic Multi-Granularity Semantics with Caption-Guided Enhancement for Few-Shot Learning

arXiv 2601.14111 ↗First posted 20 Jan 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Few-shot learning aims to identify novel categories from only a handful of labeled samples, where prototypes estimated from scarce data are often biased and generalize poorly. Semantic-based methods alleviate this by introducing coarse class-level information, but they are mostly applied on the support side, leaving query representations unchanged. In this paper, we present PMCE, a Probabilistic few-shot framework that leverages Multi-granularity semantics with Caption-guided Enhancement. PMCE constructs a nonparametric knowledge bank that stores visual statistics for each category as well as CLIP-encoded class name embeddings of the base classes. At meta-test time, the most relevant base classes are retrieved based on the similarities of class name embeddings for each novel category. These statistics are then aggregated into category-specific prior information and fused with the support set prototypes via a simple MAP update. Simultaneously, a frozen BLIP captioner provides label-free instance-level image descriptions, and a lightweight enhancer trained on base classes optimizes both support prototypes and query features under an inductive protocol with a consistency regularization to stabilize noisy captions. Experiments on four benchmarks show that PMCE consistently improves over strong baselines, achieving up to 7.71% absolute gain over the strongest semantic competitor on MiniImageNet in the 1-shot setting. Our code is available at https://anonymous.4open.science/r/PMCE-275D

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 47 of its 77 references that are in the map. First posted on arXiv 20 Jan 2026.

Builds on

47 of 77 references.

References · 47

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  4. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  5. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  6. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  7. 2022BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generationcited by 4,310
  8. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  9. 2016Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotationscited by 2,646
  10. 2016Matching Networks for One Shot Learningcited by 2,246
  11. 2021LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairscited by 1,544
  12. 2021Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Conceptscited by 1,209
35 more
  1. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,139
  2. 2019A Closer Look at Few-shot Classificationcited by 679
  3. 2019Meta-Learning with Differentiable Convex Optimizationcited by 461
  4. 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
  5. 2020Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?cited by 435
  6. 2018TADAM: Task dependent adaptive metric for improved few-shot learningcited by 408
  7. 2018Meta-learning with differentiable closed-form solverscited by 395
  8. 2020Cross-Domain Few-Shot Classification via Learned Feature-Wise Transformationcited by 183
  9. 2021Free Lunch for Few-shot Learning: Distribution Calibrationcited by 165
  10. 2019Adaptive Cross-Modal Few-Shot Learningcited by 95
  11. 2021Semantics Disentangling for Generalized Zero-Shot Learningcited by 56
  12. 2023StyleAdv: Meta Style Adversarial Training for Cross-Domain Few-Shot Learningcited by 53
  13. 2023A Closer Look at Few-shot Classification Againcited by 46
  14. 2022Adversarial Feature Augmentation for Cross-domain Few-shot Classificationcited by 36
  15. 2022Decomposed Soft Prompt Guided Fusion Enhancing for Compositional Zero-Shot Learningcited by 36
  16. 2020A Novel Perspective to Zero-shot Learning: Towards an Alignment of Manifold Structures via Semantic Feature Expansioncited by 31
  17. 2023Semantic Prompt for Few-Shot Image Recognitioncited by 29
  18. 2024Flatten Long-Range Loss Landscapes for Cross-Domain Few-Shot Learningcited by 26
  19. 2022Matching Feature Sets for Few-Shot Image Classificationcited by 25
  20. 2023Zero-Shot Learning by Harnessing Adversarial Samplescited by 24
  21. 2022Few-Shot Learning Meets Transformer: Unified Query-Support Transformers for Few-Shot Classificationcited by 16
  22. 2021SEGA: Semantic Guided Attention on Visual Prototype for Few-Shot Learningcited by 15
  23. 2023CNS-Net: Conservative Novelty Synthesizing Network for Malware Recognition in an Open-set Scenariocited by 15
  24. 2023Simple Semantic-Aided Few-Shot Learningcited by 15
  25. 2021Sparse Spatial Transformers for Few-Shot Learningcited by 14
  26. 2022GSMFlow: Generation Shifts Mitigating Flow for Generalized Zero-Shot Learningcited by 12
  27. 2024Fine-Grained Zero-Shot Learning: Advances, Challenges, and Prospectscited by 12
  28. 2024Exploring Cross-Domain Few-Shot Classification via Frequency-Aware Promptingcited by 9
  29. 2025Connecting Giants: Synergistic Knowledge Transfer of Large Multimodal Models for Few-Shot Learningcited by 9
  30. 2023Few-Shot Learning with Visual Distribution Calibration and Cross-Modal Distribution Alignmentcited by 7
  31. 2025VT-FSL: Bridging Vision and Text with LLMs for Few-Shot Learningcited by 7
  32. 2025Cluster-Aware Prompt Ensemble Learning for Few-Shot Vision-Language Model Adaptationcited by 7
  33. 2025SVIP: Semantically Contextualized Visual Patches for Zero-Shot Learningcited by 6
  34. 2021CORL: Compositional Representation Learning for Few-Shot Classificationcited by 4
  35. 2025Slot Attention-based Feature Filtering for Few-Shot Learningcited by 1

Built on by

Not cited yet.