---
title: "Bi-CoG: Bi-Consistency-Guided Self-Training for Vision-Language Models"
authors:
  - "Rui Zhu"
  - "Song-Lin Lv"
  - "Zi-Kang Wang"
  - "Lan-Zhe Guo"
doi: "10.24963/ijcai.2026/607"
url: "https://ijcai.org/proceedings/2026/607"
pdf: "https://ijcai.org/proceedings/2026/0607.pdf"
venue: "Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence"
year: 2026
pages: "5448-5456"
track: "Main Track"
keywords:
  - "Machine Learning: Semi-supervised learning"
  - "Machine Learning: Multi-modal learning"
---

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

Rui Zhu, Song-Lin Lv, Zi-Kang Wang, Lan-Zhe Guo

*Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence*, pages 5448-5456. DOI: [10.24963/ijcai.2026/607](https://doi.org/10.24963/ijcai.2026/607)

## Abstract

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 Bi-Consistency-Guided 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.