deploy: 25760a7

_sources/tutorial.rst.txt

@@ -49,7 +49,7 @@ and multispectral data at the same time is not currently supported. Stick to
 one data type per model (or stack the RGB bands with the multispectral bands
 and treat as in the case of multispectral data).
 
-Preparing data (RGB and multispectral)
+Preparing data (RGB/multispectral)
 --------------------------------------
 
 An example of the recommended file structure when training a new model is as follows:
@@ -271,7 +271,7 @@ multispectral (``.tif``) tiles.
       display(Image.fromarray(image))
 
 
-Training a model (RGB)
+Training (RGB)
 ----------------------
 
 Before training can commence, it is necessary to register the training data. It is possible to set a validation fold for
@@ -357,7 +357,7 @@ Training outputs, including model weights and training metrics, will be stored i
    Early stopping is implemented and will be triggered by a sustained failure to improve on the performance of
    predictions on the validation fold. This is measured as the AP50 score of the validation predictions.
 
-Training a model (multispectral)
+Training (multispectral)
 --------------------------------
 
 The process for training a multispectral model is similar to that for RGB data but there are some key steps that are
@@ -547,7 +547,7 @@ model is to be used on a range of different resolutions, random resizing can hel
 different scales.
 
 
-Post-training (check training convergence)
+Post-training (check convergence)
 ------------------------------------------
 
 It is important to check that the model has converged and is not overfitting. This can be done by plotting the training

_sources/tutorial_multi.rst.txt

@@ -17,8 +17,8 @@ The key steps are:
 4. Making landscape level predictions
 
 
-Preparing data (RGB and multispectral)
---------------------------------------
+Preparing data
+--------------
 
 Data can be prepared in a similar way to the single class case but the classes
 and their order (mapping) need to be saved so that they can be accessed

index.html

@@ -106,18 +106,18 @@ <h2>Citation<a class="headerlink" href="#citation" title="Link to this heading">
 </ul>
 </li>
 <li class="toctree-l1"><a class="reference internal" href="tutorial.html">2. Tutorial</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="tutorial.html#preparing-data-rgb-and-multispectral">2.1. Preparing data (RGB and multispectral)</a></li>
-<li class="toctree-l2"><a class="reference internal" href="tutorial.html#training-a-model-rgb">2.2. Training a model (RGB)</a></li>
-<li class="toctree-l2"><a class="reference internal" href="tutorial.html#training-a-model-multispectral">2.3. Training a model (multispectral)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="tutorial.html#preparing-data-rgb-multispectral">2.1. Preparing data (RGB/multispectral)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="tutorial.html#training-rgb">2.2. Training (RGB)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="tutorial.html#training-multispectral">2.3. Training (multispectral)</a></li>
 <li class="toctree-l2"><a class="reference internal" href="tutorial.html#data-augmentation">2.4. Data augmentation</a></li>
-<li class="toctree-l2"><a class="reference internal" href="tutorial.html#post-training-check-training-convergence">2.5. Post-training (check training convergence)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="tutorial.html#post-training-check-convergence">2.5. Post-training (check convergence)</a></li>
 <li class="toctree-l2"><a class="reference internal" href="tutorial.html#performance-metrics">2.6. Performance metrics</a></li>
 <li class="toctree-l2"><a class="reference internal" href="tutorial.html#evaluating-model-performance">2.7. Evaluating model performance</a></li>
 <li class="toctree-l2"><a class="reference internal" href="tutorial.html#generating-landscape-predictions">2.8. Generating landscape predictions</a></li>
 </ul>
 </li>
 <li class="toctree-l1"><a class="reference internal" href="tutorial_multi.html">3. Tutorial (multiclass)</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="tutorial_multi.html#preparing-data-rgb-and-multispectral">3.1. Preparing data (RGB and multispectral)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="tutorial_multi.html#preparing-data">3.1. Preparing data</a></li>
 <li class="toctree-l2"><a class="reference internal" href="tutorial_multi.html#training-models">3.2. Training models</a></li>
 <li class="toctree-l2"><a class="reference internal" href="tutorial_multi.html#landscape-predictions">3.3. Landscape predictions</a></li>
 </ul>

tutorial.html

@@ -51,11 +51,11 @@
 <ul class="current">
 <li class="toctree-l1"><a class="reference internal" href="installation.html">1. Installation</a></li>
 <li class="toctree-l1 current"><a class="current reference internal" href="#">2. Tutorial</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="#preparing-data-rgb-and-multispectral">2.1. Preparing data (RGB and multispectral)</a></li>
-<li class="toctree-l2"><a class="reference internal" href="#training-a-model-rgb">2.2. Training a model (RGB)</a></li>
-<li class="toctree-l2"><a class="reference internal" href="#training-a-model-multispectral">2.3. Training a model (multispectral)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="#preparing-data-rgb-multispectral">2.1. Preparing data (RGB/multispectral)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="#training-rgb">2.2. Training (RGB)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="#training-multispectral">2.3. Training (multispectral)</a></li>
 <li class="toctree-l2"><a class="reference internal" href="#data-augmentation">2.4. Data augmentation</a></li>
-<li class="toctree-l2"><a class="reference internal" href="#post-training-check-training-convergence">2.5. Post-training (check training convergence)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="#post-training-check-convergence">2.5. Post-training (check convergence)</a></li>
 <li class="toctree-l2"><a class="reference internal" href="#performance-metrics">2.6. Performance metrics</a></li>
 <li class="toctree-l2"><a class="reference internal" href="#evaluating-model-performance">2.7. Evaluating model performance</a></li>
 <li class="toctree-l2"><a class="reference internal" href="#generating-landscape-predictions">2.8. Generating landscape predictions</a></li>
@@ -130,8 +130,8 @@ <h1><span class="section-number">2. </span>Tutorial<a class="headerlink" href="#
 and multispectral data at the same time is not currently supported. Stick to
 one data type per model (or stack the RGB bands with the multispectral bands
 and treat as in the case of multispectral data).</p>
-<section id="preparing-data-rgb-and-multispectral">
-<h2><span class="section-number">2.1. </span>Preparing data (RGB and multispectral)<a class="headerlink" href="#preparing-data-rgb-and-multispectral" title="Link to this heading"></a></h2>
+<section id="preparing-data-rgb-multispectral">
+<h2><span class="section-number">2.1. </span>Preparing data (RGB/multispectral)<a class="headerlink" href="#preparing-data-rgb-multispectral" title="Link to this heading"></a></h2>
 <p>An example of the recommended file structure when training a new model is as follows:</p>
 <div class="highlight-bash notranslate"><div class="highlight"><pre><span></span>├──<span class="w"> </span>Danum<span class="w">                                       </span><span class="o">(</span>site<span class="w"> </span>directory<span class="o">)</span>
 │<span class="w">   </span>├──<span class="w"> </span>rgb
@@ -325,8 +325,8 @@ <h2><span class="section-number">2.1. </span>Preparing data (RGB and multispectr
 </pre></div>
 </div>
 </section>
-<section id="training-a-model-rgb">
-<h2><span class="section-number">2.2. </span>Training a model (RGB)<a class="headerlink" href="#training-a-model-rgb" title="Link to this heading"></a></h2>
+<section id="training-rgb">
+<h2><span class="section-number">2.2. </span>Training (RGB)<a class="headerlink" href="#training-rgb" title="Link to this heading"></a></h2>
 <p>Before training can commence, it is necessary to register the training data. It is possible to set a validation fold for
 model evaluation (which can be helpful for tuning models). The validation fold can be changed over different training
 steps to expose the model to the full range of available training data. Register as many different folders as necessary</p>
@@ -398,8 +398,8 @@ <h2><span class="section-number">2.2. </span>Training a model (RGB)<a class="hea
 predictions on the validation fold. This is measured as the AP50 score of the validation predictions.</p>
 </div>
 </section>
-<section id="training-a-model-multispectral">
-<h2><span class="section-number">2.3. </span>Training a model (multispectral)<a class="headerlink" href="#training-a-model-multispectral" title="Link to this heading"></a></h2>
+<section id="training-multispectral">
+<h2><span class="section-number">2.3. </span>Training (multispectral)<a class="headerlink" href="#training-multispectral" title="Link to this heading"></a></h2>
 <p>The process for training a multispectral model is similar to that for RGB data but there are some key steps that are
 different. Data will be read from <code class="docutils literal notranslate"><span class="pre">.tif</span></code> files of 4 or more bands instead of the 3-band <code class="docutils literal notranslate"><span class="pre">.png</span></code> files.</p>
 <p>Data should be registered as before:</p>
@@ -549,8 +549,8 @@ <h2><span class="section-number">2.4. </span>Data augmentation<a class="headerli
 model is to be used on a range of different resolutions, random resizing can help the model learn to detect objects at
 different scales.</p>
 </section>
-<section id="post-training-check-training-convergence">
-<h2><span class="section-number">2.5. </span>Post-training (check training convergence)<a class="headerlink" href="#post-training-check-training-convergence" title="Link to this heading"></a></h2>
+<section id="post-training-check-convergence">
+<h2><span class="section-number">2.5. </span>Post-training (check convergence)<a class="headerlink" href="#post-training-check-convergence" title="Link to this heading"></a></h2>
 <p>It is important to check that the model has converged and is not overfitting. This can be done by plotting the training
 and validation loss over time. The <code class="docutils literal notranslate"><span class="pre">detectron2</span></code> training routine will output a <code class="docutils literal notranslate"><span class="pre">metrics.json</span></code> file that can be used
 to plot the training and validation loss. The following code can be used to plot the loss:</p>

tutorial_multi.html

@@ -50,7 +50,7 @@
 <li class="toctree-l1"><a class="reference internal" href="installation.html">1. Installation</a></li>
 <li class="toctree-l1"><a class="reference internal" href="tutorial.html">2. Tutorial</a></li>
 <li class="toctree-l1 current"><a class="current reference internal" href="#">3. Tutorial (multiclass)</a><ul>
-<li class="toctree-l2"><a class="reference internal" href="#preparing-data-rgb-and-multispectral">3.1. Preparing data (RGB and multispectral)</a></li>
+<li class="toctree-l2"><a class="reference internal" href="#preparing-data">3.1. Preparing data</a></li>
 <li class="toctree-l2"><a class="reference internal" href="#training-models">3.2. Training models</a></li>
 <li class="toctree-l2"><a class="reference internal" href="#landscape-predictions">3.3. Landscape predictions</a></li>
 </ul>
@@ -99,8 +99,8 @@ <h1><span class="section-number">3. </span>Tutorial (multiclass)<a class="header
 <li><p>Evaluating model performance</p></li>
 <li><p>Making landscape level predictions</p></li>
 </ol>
-<section id="preparing-data-rgb-and-multispectral">
-<h2><span class="section-number">3.1. </span>Preparing data (RGB and multispectral)<a class="headerlink" href="#preparing-data-rgb-and-multispectral" title="Link to this heading"></a></h2>
+<section id="preparing-data">
+<h2><span class="section-number">3.1. </span>Preparing data<a class="headerlink" href="#preparing-data" title="Link to this heading"></a></h2>
 <p>Data can be prepared in a similar way to the single class case but the classes
 and their order (mapping) need to be saved so that they can be accessed
 consistently across training and prediction. The classes are saved in a json