Skip to content

generate caption for image using local model, Describe the image objects browsed by NVDA, support shortcut key recognition.Result will copy to clipboard automatically .

License

Notifications You must be signed in to change notification settings

tianzeshi-study/CaptionLocal

Repository files navigation

CaptionLocal

generate caption for image using local model, Describe the image objects browsed by NVDA, support shortcut key recognition.Result will copy to clipboard automatically .

使用本地模型生成图像的描述文字,描述由NVDA浏览到的图像对象。支持快捷键识别。结果将自动复制到剪贴板。

usage 用法

First install this add-on, then open model manager to download models. When models are ready, you can recognize the image navigated by NVDA (default focused object) using shortcut: NVDA+windows+,

首先安装此附加组件,然后打开模型管理器以下载模型。准备就绪时,您可以使用以下快捷方式识别NVDA对象导航到的图像(默认聚焦到的对象):NVDA+Windows+,

shortcut 快捷键 function 功能
NVDA+windows+, directly describe image objects browsed by NVDA using local model.直接使用本地模型描述浏览到的对象上的图片
NVDA+windows+shift+, release model and free memory useage. 释放模型所占用的内存
NVDA+windows+control+, open model manager to download models.打开模型管理器以便手动下载模型

model download 模型下载

use shortcut(NVDA+windows+control+,) open model manager, press tab to find download(下载) button, click it .

使用快捷方式(nvda+Windows+Control+)打开模型管理器,通过多次点击tab,首先找到高级选项, 点开后, 选上使用镜像的复选框,点击确定。回到主页面通过tab键定位到下载按钮,点击它,然后耐心等待下载完成。

About

generate caption for image using local model, Describe the image objects browsed by NVDA, support shortcut key recognition.Result will copy to clipboard automatically .

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages