Segment unknown object match text description by generative and discriminative models with Ref-Diff
Segment unknown object match text description by generative and discriminative models with Ref-Diff
Ref-Diff: Zero-shot Referring Image Segmentation with Generative Models
arXiv paper abstract https://arxiv.org/abs/2308.16777
arXiv PDF paper https://arxiv.org/pdf/2308.16777.pdf
Zero-shot referring image segmentation ... aims to find an instance segmentation mask based on the given referring descriptions, without training on this type of paired data.
Current ... methods mainly ... using pre-trained discriminative models (e.g., CLIP). However... generative models (e.g., Stable Diffusion) have potentially understood the relationships between various visual elements and text descriptions
... introduce a novel Referring Diffusional segmentor (Ref-Diff) for this task, which leverages the fine-grained multi-modal information from generative models.
... demonstrate that without a proposal generator, a generative model alone can achieve comparable performance to existing SOTA weakly-supervised models.
When ... combine both generative and discriminative models, ... Ref-Diff outperforms these competing methods by a significant margin.
This indicates that generative models are also beneficial for this task and can complement discriminative models for better referring segmentation ...
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Comments