mlsanityData to 13 Oct 2026

Bi-CoG: Bi-Consistency-Guided Self-Training for Vision-Language Models

arXiv 2510.20477 ↗First posted 23 Oct 2025cs.LGcited by 0

On the map: Multimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP

also in 1 more
Abstract (arXiv)

Exploiting unlabeled data through semi-supervised learning (SSL) or leveraging pre-trained models via fine-tuning are two prevailing paradigms for addressing label-scarce scenarios. Recently, growing attention has been given to combining fine-tuning of pre-trained vision-language models (VLMs) with SSL, forming the emerging paradigm of semi-supervised fine-tuning. However, existing methods often suffer from model bias and hyperparameter sensitivity, due to reliance on prediction consistency or pre-defined confidence thresholds. To address these limitations, we propose a simple yet effective plug-and-play methodology named underlineBi-Consistency-underlineGuided Self-Training (Bi-CoG), which assigns high-quality and low-bias pseudo-labels, by simultaneously exploiting inter-model and intra-model consistency, along with an error-aware dynamic pseudo-label assignment strategy. Both theoretical analysis and extensive experiments over 14 datasets demonstrate the effectiveness of Bi-CoG, which consistently and significantly improves the performance of existing methods.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Placed in Prompt learning and adaptation of CLIP, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 33 of its 52 references that are in the map. First posted on arXiv 23 Oct 2025.

Builds on

33 of 52 references.

Closest ancestors · 12

Same direction, newest first

  1. 2025Vision-Language Model Selection and Reuse for Downstream Adaptation30 Jan
  2. 2025BMIP: Bi-directional Modality Interaction Prompt Learning for VLM14 Jan
  3. 2024Vision-Language Model Fine-Tuning via Simple Parameter-Efficient Modification25 Sep
  4. 2024Candidate Pseudolabel Learning: Enhancing Vision-Language Models by Prompt Tuning with Unlabeled Data15 Jun
  5. 2024DeCoOp: Robust Prompt Tuning with Out-of-Distribution Detection1 Jun
  6. 2024Revisiting Active Learning in the Era of Vision Foundation Models25 Jan
  7. 2023Enhancing CLIP with CLIP: Exploring Pseudolabeling for Limited-Label Prompt Tuning2 Jun
  8. 2022Tip-Adapter: Training-free Adaption of CLIP for Few-shot Classification19 Jul
  9. 2022Unsupervised Prompt Learning for Vision-Language Models7 Apr
  10. 2022Conditional Prompt Learning for Vision-Language Models10 Mar
  11. 2021CLIP-Adapter: Better Vision-Language Models with Feature Adapters9 Oct
  12. 2021Learning to Prompt for Vision-Language Models2 Sep

Other references · 21

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,588
  2. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,855
  3. 2013Describing Textures in the Wildcited by 2,115
  4. 2020FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidencecited by 2,082
  5. 2013Fine-Grained Visual Classification of Aircraftcited by 1,709
  6. 2017EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classificationcited by 1,492
  7. 2022MaPLe: Multi-modal Prompt Learningcited by 685
  8. 2017Skin Lesion Analysis Toward Melanoma Detection: A Challenge at the 2017 International Symposium on Biomedical Imaging (ISBI), Hosted by the International Skin Imaging Collaboration (ISIC)cited by 549
  9. 2021FlexMatch: Boosting Semi-Supervised Learning with Curriculum Pseudo Labelingcited by 539
  10. 2023Self-regulating Prompts: Foundational Model Adaptation without Forgettingcited by 272
  11. 2021A Survey on Deep Semi-supervised Learningcited by 250
  12. 2020CoMatch: Semi-supervised Learning with Contrastive Graph Regularizationcited by 167
9 more
  1. 2022Debiased Learning from Naturally Imbalanced Pseudo-Labelscited by 85
  2. 2022Debiased Self-Training for Semi-Supervised Learningcited by 62
  3. 2023Overwriting Pretrained Bias with Finetuning Datacited by 32
  4. 2021The Rich Get Richer: Disparate Impact of Semi-Supervised Learningcited by 28
  5. 2024Erasing the Bias: Fine-Tuning Foundation Models for Semi-Supervised Learningcited by 19
  6. 2025Neuro-Symbolic Artificial Intelligence: Towards Improving the Reasoning Abilities of Large Language Modelscited by 19
  7. 2024ChinaTravel: An Open-Ended Travel Planning Benchmark with Compositional Constraint Validation for Language Agentscited by 15
  8. 2024Robust Semi-Supervised Learning in Open Environmentscited by 8
  9. 2025Unlabeled Data vs. Pre-trained Knowledge: Rethinking SSL in the Era of Large Modelscited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025SelfPrompt: Confidence-Aware Semi-Supervised Tuning for Robust Vision-Language Model Adaptation
  2. 2026Fine-tuning Pre-trained Vision-Language Models in a Human-Annotation-Free Manner
  3. 2025Test-Time Consistency in Vision Language Models
  4. 2024Pre-Trained Vision-Language Models as Partial Annotators
  5. 2025Bridging Weakly-Supervised Learning and VLM Distillation: Noisy Partial Label Learning for Efficient Downstream Adaptation
  6. 2026SemAlign: Language Guided Semi-supervised Domain Generalization
  7. 2026Semi-Supervised Few-Shot Adaptation of Vision-Language Models
  8. 2022Training Vision-Language Models with Less Bimodal Supervision