Symmetrical Visual Contrastive Optimization: Aligning Vision-Language Models with Minimal Contrastive Images
Abstract
Recent studies have shown that Large Vision-Language Models (VLMs) tend to neglect image content and over-rely on language-model priors, resulting in errors in visually grounded tasks and hallucinations. We hypothesize that this issue arises because existing VLMs are not explicitly trained to generate texts that are accurately grounded in fine-grained image details. To enhance visual feedback during VLM training, we propose S-VCO (Symmetrical Visual Contrastive Optimization), a novel finetuning objective that steers the model toward capturing important visual details and aligning them with corresponding text tokens. To further facilitate this detailed alignment, we introduce MVC, a paired image-text dataset built by automatically filtering and augmenting visual counterfactual data to challenge the model with hard contrastive cases involving Minimal Visual Contrasts. Experiments show that our method consistently improves VLM performance across diverse benchmarks covering various abilities and domains, achieving up to a 22% reduction in hallucinations, and significant gains in vision-centric and general tasks. Notably, these improvements become increasingly pronounced in benchmarks with higher visual dependency. In short, S-VCO offers a significant enhancement of VLM's visually-dependent task performance while retaining or even improving the model's general abilities. We opensource our code at https://s-vco.github.io/
Community
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
- Re-Align: Aligning Vision Language Models via Retrieval-Augmented Direct Preference Optimization (2025)
- Probing Visual Language Priors in VLMs (2024)
- Visual Attention Never Fades: Selective Progressive Attention ReCalibration for Detailed Image Captioning in Multimodal Large Language Models (2025)
- Why Vision Language Models Struggle with Visual Arithmetic? Towards Enhanced Chart and Geometry Understanding (2025)
- Supervision-free Vision-Language Alignment (2025)
- ImageRef-VL: Enabling Contextual Image Referencing in Vision-Language Models (2025)
- SMIR: Efficient Synthetic Data Pipeline To Improve Multi-Image Reasoning (2025)
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment:
@librarian-bot
recommend
Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper
Collections including this paper 0
No Collection including this paper