|
<!DOCTYPE html> |
|
<html lang="en"> |
|
<head> |
|
<meta charset="UTF-8"> |
|
<meta name="viewport" content="width=device-width, initial-scale=1.0"> |
|
<title>Experiment Interface Guide</title> |
|
|
|
<link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.0.0-beta3/css/all.min.css" rel="stylesheet"> |
|
<style> |
|
body { |
|
font-family: Arial, sans-serif; |
|
line-height: 1.6; |
|
margin: 20px; |
|
} |
|
h2 { |
|
color: #2c3e50; |
|
} |
|
.section { |
|
margin-bottom: 20px; |
|
} |
|
ul { |
|
list-style-type: disc; |
|
margin-left: 20px; |
|
} |
|
.icon { |
|
margin-left: 5px; |
|
color: #3498db; |
|
} |
|
|
|
@media (prefers-color-scheme: dark) { |
|
.highlight { |
|
display: inline-block; |
|
background-color: #333; |
|
color: #ffffff !important; |
|
padding: 2px 5px; |
|
border-radius: 5px; |
|
border: 1px solid rgba(0, 0, 0, 0.2); |
|
} |
|
.button { |
|
display: inline-block; |
|
background-color: #f97316; |
|
color: #f39c12 !important; |
|
padding: 2px 5px; |
|
border-radius: 5px; |
|
border: 1px solid rgba(0, 0, 0, 0.2); |
|
} |
|
} |
|
|
|
@media (prefers-color-scheme: light) { |
|
.highlight { |
|
display: inline-block; |
|
background-color: #ffe8ef; |
|
color: #C01343 !important; |
|
padding: 2px 5px; |
|
border-radius: 5px; |
|
border: 1px solid rgba(0, 0, 0, 0.2); |
|
} |
|
.button { |
|
display: inline-block; |
|
background-color: #fdba74; |
|
color: #ea580c !important; |
|
padding: 2px 5px; |
|
border-radius: 5px; |
|
border: 1px solid rgba(0, 0, 0, 0.2); |
|
} |
|
} |
|
|
|
</style> |
|
</head> |
|
<body> |
|
<div class="section"> |
|
<h1>Detection</h1> |
|
<p>This tab displays the structure of the model inputted into the generated <span class="highlight"> Experiment </span>. It shows a directed graph where each layer is represented as a node and each connection as an edge.</p> |
|
</div> |
|
|
|
<div class="section"> |
|
<h1>Local Explanation</h1> |
|
<p>This tab allows users to select a local instance for the generated <span class="highlight"> Experiment </span>, automatically generate explanations, and perform evaluation and validation. It automatically classifies the available XAI (Explainable AI) methods for the input model and provides a UI for users to select the necessary methods.</p> |
|
<p>The usage steps are as follows:</p> |
|
<ul> |
|
<li>Select an input from the <span class="highlight"> Input Data Gallery </span>. </li> |
|
<li>In the <span class="highlight">Explainers</span> section, open the dropdown to select a <span class="highlight"> Default parameter </span> or click the <span class="highlight"> Optimize </span> button to perform hyperparameter optimization, and then select the <span class="button"> Optimized parameter </span> to use the corresponding XAI method.</li> |
|
<li>In the <span class="highlight">Evaluators</span> section, choose a metric to evaluate the XAI.</li> |
|
<li>Click the <span class="button">Explain</span> button. </li> |
|
</ul> |
|
<p>The explanation results for the selected XAI methods will be displayed and sorted in order of the selected metric's results.</p> |
|
<p><em>*Demo server is fixed to use 'AbPC' metric as an object of the optimization and ranking.</em></p> |
|
</div> |
|
|
|
</body> |
|
</html> |
|
|
