Say Anything: Prompt Segment Anything Model

Say Anything: Prompt Segment Anything Model

5|0 评价|0 收藏
访问网站
产品介绍
Enable natural language prompting for the Segment Anything Model (SAM) using CLIP embeddings.
收录时间
2026年7月22日
免费浏览器插件AI 图像标注AI 图像分类AI 图像分割AI 物体检测AI 场景理解

什么是 Say Anything: Prompt Segment Anything Model？

Say Anything is a Chrome extension that enables natural language prompting for Meta's Segment Anything Model (SAM). It uses OpenAI's CLIP and kevinzakka's implementation of Gradient-weighted Class Activation Mapping to allow users to interact with SAM using text prompts. The extension is a layer built on top of SAM's encoding, providing a fun and experimental way to segment images with natural language.

如何使用 Say Anything: Prompt Segment Anything Model？

  • Install the extension from the Chrome Web Store. Visit the GitHub repository (https://github.com/suvansh/say-anything-backend) to set up the backend server locally in 2 minutes. Use natural language prompts to interact with the Segment Anything Model.

Say Anything: Prompt Segment Anything Model 的核心功能

  • Natural language prompting for Segment Anything Model (SAM)
  • Uses CLIP embeddings for text-based segmentation
  • Built on top of SAM with Gradient-weighted Class Activation Mapping
  • Open-source backend setup available on GitHub
  • Lightweight extension (70.72KiB)

用户评价

5.0

/ 5

0 条评价

5 星
0
4 星
0
3 星
0
2 星
0
1 星
0

暂无评价，来写第一条吧

相似产品

Say Anything: Prompt Segment Anything Model 推广组件

使用网站徽章展示您的产品已收录于 Hootool，帮助更多人按问题找到合适的 AI 工具。推广组件可轻松添加到主页或页脚。

Featured on

Hootool.ai