lasa qualitative

docs/index.html

@@ -218,41 +218,78 @@ <h2 class="title is-3">Video Presentation</h2>
   <!-- End youtube video -->
 
 
+  <!-- Experiment 1 carousel -->
   <!-- Experiment 1 carousel -->
   <section class="hero is-small is-light">
     <div class="hero-body">
       <div class="container">
         <h2 class="title is-3">Experiment 1: Imitating 2D Trajectories</h2>
+
+        <!-- Pipeline overview -->
         <div class="hero-body">
           <h2 class="subtitle has-text-centered">
             Experiment 1 Pipeline overview
           </h2>
           <img src="static/images/exp_1.png" alt="ANDPs Pipeline" style="width: 50%; margin-left: 25%;">
-          <div id="results-carousel" class="carousel results-carousel">
-            <div class="item">
-
-              <h2 class="subtitle has-text-centered">
-                Qualitative results for ANDPs, SEDS, and NNs in 4 selected tasks of the LASA dataset.
-              </h2>
-              <img src="static/images/lasa_qualitative.png" alt="comparisons lasa qualitative" />
-
-            </div>
-            <div class="item">
+        </div>
 
-              <h2 class="subtitle has-text-centered">
-                Training and Test set errors on selected tasks for the quantitative comparison of ANDPs with SEDS and
-                NNs.
-              </h2>
-              <img src="static/images/train_errors_final.png" alt="errors lasa quantitative" height="50%" />
+        <!-- First Carousel: Qualitative and Quantitative Results -->
+        <div id="results-carousel-1" class="carousel results-carousel">
+          <div class="item">
+            <h2 class="subtitle has-text-centered">
+              Qualitative results for ANDPs, SEDS, and NNs in 4 selected tasks of the LASA dataset.
+            </h2>
+            <img src="static/images/lasa_qualitative.png" alt="comparisons lasa qualitative" />
+          </div>
+          <div class="item">
+            <h2 class="subtitle has-text-centered">
+              Training and Test set errors on selected tasks for the quantitative comparison of ANDPs with SEDS and NNs.
+            </h2>
+            <img src="static/images/train_errors_final.png" alt="errors lasa quantitative"
+              style="width: 80%; margin-left: 10%;" />
+          </div>
+        </div>
 
-            </div>
+        <!-- Second Carousel: Qualitative Results in Selected Splits -->
+        <div id="results-carousel-2" class="carousel results-carousel">
+          <div class="item">
+            <h2 class="subtitle has-text-centered">
+              Qualitative results for ANDPs, SEDS, and NNs in selected splits of the LASA dataset.
+            </h2>
+            <img src="static/images/lasa_qualitative_0.png" alt="Lasa Qualitative 1/5" style="width: 80%; margin-left: 10%;">
+          </div>
+          <div class="item">
+            <h2 class="subtitle has-text-centered">
+              Qualitative results for ANDPs, SEDS, and NNs in selected splits of the LASA dataset.
+            </h2>
+            <img src="static/images/lasa_qualitative_1.png" alt="Lasa Qualitative 2/5" style="width: 80%; margin-left: 10%;">
+          </div>
+          <div class="item">
+            <h2 class="subtitle has-text-centered">
+              Qualitative results for ANDPs, SEDS, and NNs in selected splits of the LASA dataset.
+            </h2>
+            <img src="static/images/lasa_qualitative_2.png" alt="Lasa Qualitative 3/5" style="width: 80%; margin-left: 10%;">
           </div>
+        <div class="item">
+          <h2 class="subtitle has-text-centered">
+            Qualitative results for ANDPs, SEDS, and NNs in selected splits of the LASA dataset.
+          </h2>
+          <img src="static/images/lasa_qualitative_3.png" alt="Lasa Qualitative 4/5" style="width: 80%; margin-left: 10%;">
         </div>
+      <div class="item">
+        <h2 class="subtitle has-text-centered">
+          Qualitative results for ANDPs, SEDS, and NNs in selected splits of the LASA dataset.
+        </h2>
+        <img src="static/images/lasa_qualitative_4.png" alt="ANDPs Lasa Qualitative 5/5" style="width: 80%; margin-left: 10%;">
+      </div>
+    </div>
       </div>
     </div>
   </section>
   <!-- End Experiment 1 carousel -->
 
+  <!-- End Experiment 1 carousel -->
+
   <!-- Experiment 2 carousels -->
   <section class="hero is-small">
     <div class="hero-body">
@@ -486,22 +523,29 @@ <h2 class="subtitle has-text-centered">
       <div class="container">
         <h2 class="title is-3">Supplementary Experiment: Learning a spiral motion in joint space</h2>
         <p>In order to showcase that ANDPs can work with arbitrary state spaces
-          <sup><a href="#footnote1">1</a></sup>, we learn the spiral motion but this time in joint space (7 degrees of freedom), this means that <strong><i>x</i></strong> ≡ <strong><i>x<sub>c</sub></i></strong> =
-              <span style="display:inline-block; vertical-align:middle;">
-                  <table style="display:inline-table;">
-                      <tr>
-                          <td>θ<sub>0</sub></td>
-                          <td>θ<sub>1</sub></td>
-                          <td>⋯</td>
-                          <td>θ<sub>6</sub></td>
-                      </tr>
-                  </table>
-              </span>
-              and the output is the desired velocity profile <strong>ẋ<sub>c</sub></strong> that the joints should follow.
-          The data collection process of this experiment is similar to the one described above, with the only difference that this time on every timestep we collect the angle of every joint of the robot, as well as their corresponding angular velocities. The results showcase that we are able to learn the task even in the higher dimensional space of the joints.</p>
-
-          <hr>
-          <p id="footnote1"><sup>1</sup> Formally, we should require that the controllable part of the state forms a Euclidean space, which holds for the reduced coordinates system (joint space). The derivations and discussion on this topic are out of the scope of this paper, and will be discussed in future work.</p>
+          <sup><a href="#footnote1">1</a></sup>, we learn the spiral motion but this time in joint space (7 degrees of
+          freedom), this means that <strong><i>x</i></strong> ≡ <strong><i>x<sub>c</sub></i></strong> =
+          <span style="display:inline-block; vertical-align:middle;">
+            <table style="display:inline-table;">
+              <tr>
+                <td>θ<sub>0</sub></td>
+                <td>θ<sub>1</sub></td>
+                <td>⋯</td>
+                <td>θ<sub>6</sub></td>
+              </tr>
+            </table>
+          </span>
+          and the output is the desired velocity profile <strong>ẋ<sub>c</sub></strong> that the joints should follow.
+          The data collection process of this experiment is similar to the one described above, with the only difference
+          that this time on every timestep we collect the angle of every joint of the robot, as well as their
+          corresponding angular velocities. The results showcase that we are able to learn the task even in the higher
+          dimensional space of the joints.
+        </p>
+
+        <hr>
+        <p id="footnote1"><sup>1</sup> Formally, we should require that the controllable part of the state forms a
+          Euclidean space, which holds for the reduced coordinates system (joint space). The derivations and discussion
+          on this topic are out of the scope of this paper, and will be discussed in future work.</p>
 
       </div>