ReCalMatch:Reliability-Calibrated Semantic Guidance for Semi-Supervised Fine-Grained Recognition
cs.CV, cs.AI
Submitted: 2026-08-30
Updated: 2026-08-30
Comments: Accepted for publication at the British Machine Vision Conference (BMVC) 2026. Official list of accepted papers:https://bmvc2026.bmva.org/programme/accepted_papers/
License: http://creativecommons.org/licenses/by/4.0/
The gist: Semi-supervised fine-grained visual recognition is highly vulnerable to overconfident pseudo-label errors: visually similar categories frequently produce high-confidence yet incorrect predictions,
Terminology
Abstract
Semi-supervised fine-grained visual recognition is highly vulnerable to overconfident pseudo-label errors: visually similar categories frequently produce high-confidence yet incorrect predictions, and consistency regularization then reinforces these errors throughout training. Existing semi-supervised learning (SSL) methods estimate pseudo-label reliability almost entirely from the visual classifier itself---maximum probability, adaptive thresholds, or entropy---signals that remain blind to whether a predicted class is semantically compatible with the visual representation. We propose ReCalMatch, a reliability-calibrated semantic framework for semi-supervised fine-grained recognition. Rather than treating textual semantics as auxiliary supervision, ReCalMatch uses multi-aspect semantic prototypes as calibration evidence for pseudo-label learning. We construct class-conditioned semantic prototypes from class names and domain-specific semantic aspects, and measure a visual--semantic agreement score between each unlabeled embedding and its pseudo-label prototype. This agreement is combined with prediction confidence and entropy into a single reliability weight that down-weights pseudo-labels that are visually confident but semantically inconsistent. A semantic consistency term and a semantic margin regularizer further sharpen prototype separability under limited labels. Extensive experiments on CUB-200-2011, Stanford Dogs, NABirds, and iNaturalist18 show that ReCalMatch consistently improves strong SSL baselines, with the largest gains in low-label regimes where pseudo-label noise is most severe.
Sources
- ReMixMatch: Semi-Supervised Learning with Distribution Alignment and Augmentation Anchoring
- AdaMatch: A Unified Approach to Semi-Supervised Learning and Domain Adaptation
- SoftMatch: Addressing the Quantity-Quality Trade-off in Semi-supervised Learning
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Context-Based Semantic-Aware Alignment for Semi-Supervised Multi-Label Learning
- Revisiting Chain-of-Thought Reasoning under Limited Supervision: Semi-supervised Chain-of-Thought Learning
- Semi-Supervised Vision-Language-Action Model
- Weakly Supervised Bilinear Attention Network for Fine-Grained Visual Classification
- SelfMatch: Combining Contrastive Self-Supervision and Consistency for Semi-Supervised Learning
- Temporal Ensembling for Semi-Supervised Learning
- Triple Expert Learning from Noisy Labels for Semi-Supervised Vision Foundation Model Adaptation
- Multi-Grained Cross-modal Alignment for Learning Open-vocabulary Semantic Segmentation from Text Supervision
- Feature Fusion Vision Transformer for Fine-Grained Visual Categorization
- FreeMatch: Self-adaptive Thresholding for Semi-supervised Learning
- USB: A Unified Semi-supervised Learning Benchmark for Classification
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models