{update] 完善内容，修改导航栏侧边栏的bug

scu-zjz · Jul 3, 2024 · 5d81103 · 5d81103
1 parent 539c4c5
commit 5d81103
Show file tree

Hide file tree

Showing 10 changed files with 140 additions and 112 deletions.
diff --git a/README.md b/README.md
@@ -1,4 +1,11 @@
 # Documentation for IMDL-BenCo
 
 ## Introduction
-This repo is the documentation for [IMDL-BenCo](https://github.com/scu-zjz/IMDLBenCo) repository, the first comprehensive IMDL benchmark and modular codebase.
+This repo is the documentation for [IMDL-BenCo](https://github.com/scu-zjz/IMDLBenCo) repository, the first comprehensive IMDL benchmark and modular codebase.
+
+
+## Local Test for developers
+First, you need to install NodeJS, then use this command to host a local server to test the website.
+```
+npm run docs:dev
+```
diff --git a/docs/.vuepress/configs/navbar/en.ts b/docs/.vuepress/configs/navbar/en.ts
@@ -2,61 +2,62 @@ import type { NavbarConfig } from '@vuepress/theme-default'
 import { version } from '../meta.js'
 
 export const navbarEn: NavbarConfig = [
-  {
-      text: 'Guide',
-      children: [
-          {
-              text: 'Introduction',
-              children: [
-                  "/guide/base/introduction.md"
-              ],
-          },
-          {
-              text: 'Quick Start',
-              children: [
-                  "/guide/quickstart/install.md"
-              ],
-          },
-      ],
-  },
+    {
+        text: 'Guide',
+        children: [
+            {
+                text: 'Basic Information',
+                children: [
+                    "/guide/base/introduction.md",
+                    "/guide/base/framework.md",
+                ],
+            },
+            {
+                text: 'Quick Start',
+                children: [
+                    "/guide/quickstart/install.md"
+                ],
+            },
+        ],
+    },
 
-  {
-      text: 'API Reference',
-      children: [
-          "/API/intro.md",
-      ],
-  },
+    {
+        text: 'API Reference',
+        children: [
+            "/API/intro.md",
+        ],
+    },
 
-  {
-      text: 'Models & modules',
+    {
+        text: 'Models & modules',
 
 
-      children: [
-          {
-              text: 'Model zoo',
-              link: '/model/model_zoo/intro&content.md',
-              children: [
-                  // '/zh/model/model_zoo/intro&content.md',
-                  '/model/model_zoo/leaderboard.md',
+        children: [
+            {
+                text: 'Model zoo',
+                link: '/model/model_zoo/intro&content.md',
+                children: [
+                    // '/zh/model/model_zoo/intro&content.md',
+                    '/model/model_zoo/leaderboard.md',
 
-              ],
-          },
-          {
-              text: 'Backbone models',
-              link: '/model/backbone/intro&content.md',
-              children: [
-                  // '/zh/model/backbone/intro&content.md',
-              ],
-          },
-          {
-              text: 'Extractor modules',
-              link: '/model/extractor/intro&content.md',
-              children: [
-                  // '/zh/model/extractor/intro&content.md',
-                  // '/zh/reference/bundler/webpack.md',
-              ],
-          },
-      ],
-  },
+                ],
+            },
+            {
+                text: 'Backbone models',
+                link: '/model/backbone/intro&content.md',
+                children: [
+                    // '/zh/model/backbone/intro&content.md',
+                ],
+            },
+            {
+                text: 'Extractor modules',
+                link: '/model/extractor/intro&content.md',
+                children: [
+                    // '/zh/model/extractor/intro&content.md',
+                    // '/zh/reference/bundler/webpack.md',
+                ],
+            },
+        ],
+    },
 
 ]
diff --git a/docs/.vuepress/configs/navbar/zh.ts b/docs/.vuepress/configs/navbar/zh.ts
@@ -9,7 +9,8 @@ export const navbarZh: NavbarConfig = [
             {
                 text: '基本信息',
                 children: [
-                    "/zh/guide/base/introduction.md"
+                    "/zh/guide/base/introduction.md",
+                    "/zh/guide/base/framework.md",
                 ],
             },
             {

diff --git a/docs/.vuepress/configs/sidebar/en.ts b/docs/.vuepress/configs/sidebar/en.ts
@@ -6,7 +6,7 @@ export const sidebarEn: SidebarConfig = {
             text: 'Guide',
             children: [
                 '/guide/base/introduction.md',
-                // '/guide/getting-started.md',
+                '/guide/base/framework.md',
                 // '/guide/configuration.md',
                 // '/guide/page.md',
                 // '/guide/markdown.md',

diff --git a/docs/.vuepress/configs/sidebar/zh.ts b/docs/.vuepress/configs/sidebar/zh.ts
@@ -6,6 +6,7 @@ export const sidebarZh: SidebarConfig = {
             text: '指南',
             children: [
                 '/zh/guide/base/introduction.md',
+                '/zh/guide/base/framework.md',
                 // '/zh/guide/getting-started.md',
                 // '/zh/guide/configuration.md',
                 // '/zh/guide/page.md',

diff --git a/docs/guide/base/framework.md b/docs/guide/base/framework.md
@@ -0,0 +1,31 @@
+# Framework Design
+
+## Overview
+
+The overview of the IMDL-BenCo framework design is shown below:
+
+![IMDL-BenCo Overview](/images/IMDLBenCo_overview.png)
+
+The main components include:
+- `Dataloader` for data loading and preprocessing
+- `Model Zoo` for managing all models and feature extractors
+- GPU-accelerated `Evaluator` for calculating evaluation metrics
+
+These classes are the most carefully designed parts of the framework and can be considered the main contributions of IMDL-BenCo.
+
+Additionally, there are auxiliary components, including:
+- `Data Library` and `Data Manager` for dataset download and management (TODO)
+- Global registration mechanism `Register` for mapping `str` to specific `class` or `method`, making it easy to call corresponding models or methods via shell scripts for batch experiments.
+- `Visualize tools` for visualization analysis, currently only including Tensorboard.
+
+And some miscellaneous tools, including:
+- `PyTorch optimize tools`, mainly for PyTorch training-related interfaces and tools.
+- `Analysis tools`, mainly for various tools used for training or post-training analysis and archiving.
+
+Each of the aforementioned components is independently designed as a class or function, with appropriate interfaces for interaction between components. Ultimately, they fulfill their respective roles by being imported, called, and combined in various ·、`Training/Testing/Visualizing Scripts`.
+
+The CLI (Command Line Interface) of the entire IMDL-BenCo framework, similar to `git init` in Git, automatically generates all default `Training/Testing/Visualizing Scripts` scripts in an appropriate working path via `benco init`, for researchers to modify and use subsequently.
+
+Therefore, users are especially encouraged to modify the content of `Training/Testing/Visualizing Scripts` as needed, making reasonable use of the framework's functions to meet customization needs. According to the ❄️ and 🔥 symbols in the diagram, users are advised to create new classes or modify and design corresponding functions as needed to accomplish specific research tasks.
+
+Additionally, functions like dataset download and model checkpoint download are also achieved through CLI commands like `benco data`.
diff --git a/docs/guide/base/introduction.md b/docs/guide/base/introduction.md
@@ -1,5 +1,5 @@
 # Introduction
-
+## Overview
 IMDLBenCo is a modular deep learning toolkit designed for image manipulation detection and localization (IMDL) tasks based on the [PyTorch](https://pytorch.org/) framework. It can be used for:
 - Reproducing state-of-the-art (SoTA) models in IMDL
 - Creating your own IMDL models
@@ -32,35 +32,6 @@ Under this philosophy, IMDL-BenCo's framework has the following features and adv
   - Integrated visualization tools like Tensorboard, only the input of images and scalars into specified interfaces are required.
   - Integrated complexity analysis (parameters, FLOPs), Grad-CAM, and other analysis tools for quick and easy completion of paper charts.
 
-## Framework Design
-
-The overview of the IMDL-BenCo framework design is shown below:
-
-![IMDL-BenCo Overview](/images/IMDLBenCo_overview.png)
-
-The main components include:
-- `Dataloader` for data loading and preprocessing
-- `Model Zoo` for managing all models and feature extractors
-- GPU-accelerated `Evaluator` for calculating evaluation metrics
-
-These classes are the most carefully designed parts of the framework and can be considered the main contributions of IMDL-BenCo.
-
-Additionally, there are auxiliary tools, including:
-- `Data Library` and `Data Manager` for dataset download and management (TODO)
-- Global registration mechanism `Register` for mapping `str` to specific `class` or `method`, making it easy to call corresponding models or methods via shell scripts for batch experiments.
-- `Visualize tools` for visualization analysis, currently only including Tensorboard
-
-And some miscellaneous interfaces and tools, including:
-- `PyTorch optimize tools`, mainly for PyTorch training-related interfaces and tools.
-- `Analysis tools`, mainly for various tools used for training or post-training analysis and archiving.
-
-All the above tools form classes or functions independently, with corresponding interfaces, ultimately achieving their respective functions through various `Training/Testing/Visualizing Scripts`.
-
-The CLI (Command Line Interface) of the entire IMDL-BenCo framework, similar to `git init` in Git, automatically generates all default `Training/Testing/Visualizing Scripts` scripts in an appropriate working path via `benco init`, for researchers to modify and use subsequently.
-
-Therefore, users are especially encouraged to modify the content of `Training/Testing/Visualizing Scripts` as needed, making reasonable use of the framework's functions to meet customization needs. According to the ❄️ and 🔥 symbols in the diagram, users are advised to create new classes or modify and design corresponding functions as needed to accomplish specific research tasks.
-
-Additionally, functions like dataset download and model checkpoint download are also achieved through CLI commands like `benco data`.
 
 ## Motivation
 

diff --git a/docs/zh/guide/base/framework.md b/docs/zh/guide/base/framework.md
@@ -0,0 +1,29 @@
+# 框架设计
+## 概览
+IMDL-BenCo代码框架的设计概览图如下所示：
+
+![](/images/IMDLBenCo_overview.png)
+
+主要的组件包含：
+- 负责引入数据并进行预处理的`Dataloader`
+- 管理全部模型，特征提取器的`Model Zoo`
+- 基于GPU加速的`Evaluator`，用于计算评价指标
+
+上述类是整个框架中最精心设计的部分，可以认为是IMDL-BenCo的主要贡献。
+
+此外还有辅助的一些组件，包含：
+- 数据集下载和管理工具`Data Library`和`Data Manager`（TODO）
+- 全局的注册机制`Register`，可以实现从`str`到具体`class`或者`method`的映射，便于直接通过shell脚本调用相应的模型或方法，以便批量完成实验。
+- 用于可视化分析结果的`visualize tools`，暂时只包含Tensorboard。
+
+以及一些零碎的工具，包含：
+- `PyTorch optimize tools`，主要是PyTorch训练相关的接口和工具。
+- `Analysis tools`，主要是各种训练时或训练后，分析存档用的工具。
+
+所有上述工具，各自独立地构成了类或者函数，存在交互的组件间留有相应接口。最终，通过在多种`Training/Testing/Visualizing Scrips`中import调用并组合来实现相应的职责。
+
+而整个IMDL-BenCo框架的CLI（命令行界面，Command Line Interface）则以类似Git中`git init`的行为，通过`benco init`自动地在合适的工作路径生成所有**默认**的`Training/Testing/Visualizing Scrips`脚本，供研究人员进行后续修改使用。
+
+所以，我们尤其鼓励使用者按照需求修改`Training/Testing/Visualizing Scrips`的内容，完成对于框架功能的合理取用，满足自定义的需求。并根据图中的❄️、🔥标志建议、酌情对于标注为🔥的类按需创建新的类或修改、设计相应功能完成最相应的科研任务。
+
+此外，数据集下载，模型checkpoint下载等等功能也是通过`benco data`等等CLI指令实现。
diff --git a/docs/zh/guide/base/introduction.md b/docs/zh/guide/base/introduction.md
@@ -1,4 +1,5 @@
 # 简介
+## 概览
 IMDLBenCo 是一个模块化的，针对图像篡改检测和定位任务设计的深度学习工具包，基于的[PyTorch](https://pytorch.org/)框架设计。可以用于：
 - 复现篡改检测领域SoTA模型
 - 创建自己的篡改检测模型
@@ -30,34 +31,6 @@ IMDL-BenCo的设计理念为：尽最大可能平衡 ***科研代码对于自定
   - 集成Tensorboard等可视化工具，只需要向指定接口传入图像，标量即可。
   - 集成复杂度分析（参数量，FLOPs），Grad-CAM等分析工具，方便快捷完成论文图表。
 
-## 框架设计
-IMDL-BenCo代码框架的设计概览图如下所示：
-
-![](/images/IMDLBenCo_overview.png)
-
-主要的组件包含：
-- 负责引入数据并进行预处理的`Dataloader`
-- 管理全部模型，特征提取器的`Model Zoo`
-- 基于GPU加速的`Evaluator`，用于计算评价指标
-
-上述类是整个框架中最精心设计的部分，可以认为是IMDL-BenCo的主要贡献。
-
-此外还有辅助的一些工具，包含：
-- 数据集下载和管理工具`Data Library`和`Data Manager`（TODO）
-- 全局的注册机制`Register`，可以实现从`str`到具体`class`或者`method`的映射，便于直接通过shell脚本调用相应的模型或方法，以便批量完成实验。
-- 用于可视化分析结果的`visualize tools`，暂时只包含Tensorboard
-
-以及一些零碎的接口和工具，包含：
-- `PyTorch optimize tools`，主要是PyTorch训练相关的接口和工具。
-- `Analysis tools`，主要是各种训练时或训练后，分析存档用的工具。
-
-所有上述工具，各自独立地构成了类或者函数，并留有相应接口，最终通过多种`Training/Testing/Visualizing Scrips`调用并实现相应功能。
-
-而整个IMDL-BenCo框架的CLI（命令行界面，Command Line Interface）则以类似Git中`git init`的行为，通过`benco init`自动地在合适的工作路径生成所有**默认**的`Training/Testing/Visualizing Scrips`脚本，供研究人员进行后续修改使用。
-
-所以，我们尤其鼓励使用者按照需求修改`Training/Testing/Visualizing Scrips`的内容，完成对于框架功能的合理取用，满足自定义的需求。并根据图中的❄️、🔥标志建议、酌情对于标注为🔥的类按需创建新的类或修改、设计相应功能完成最相应的科研任务。
-
-此外，数据集下载，模型checkpoint下载等等功能也是通过`benco data`等等CLI指令实现。
 
 
 ## 动机

diff --git a/vuepress-starter/README.md b/vuepress-starter/README.md
@@ -0,0 +1,14 @@
+# Documentation for IMDL-BenCo
+
+## Introduction
+This repo is the documentation for [IMDL-BenCo](https://github.com/scu-zjz/IMDLBenCo) repository, the first comprehensive IMDL benchmark and modular codebase.
+
+
+## Local Test for developers
+First, you need to install NodeJS, then use this command to host a local server to test the website.
+```
+npm run docs:dev
+```
+Then a website server will run at http://localhost:8080/IMDLBenCo-doc/
+
+After changes to the document, the content on the webpage will change automatically.