update

index.html

@@ -87,7 +87,7 @@ <h1 class="title is-1 publication-title">Bringing Back the Context: Camera Trap
                             <a href="https://sites.google.com/view/wei-lun-harry-chao" target="_blank">Wei-Lun Chao</a><sup>1</sup>,
                             </span>
                             <span class="author-block">
-                            <a href="https://ysu1989.github.io/" target="_blank">Yu Su</a><sup>1</sup>,
+                            <a href="https://ysu1989.github.io/" target="_blank">Yu Su</a><sup>1</sup>
                             </span>
                         </div>
 
@@ -211,6 +211,7 @@ <h2 class="title is-3">Overall Results</h2>
                             <p>
                                 <b>Species Classification results on iWildCam2020-WILDS (OOD) dataset</b>. The first baseline in the second section shows the
                                 no-context baseline that uses only image-species labels as KG edges. All models use a pre-trained ResNet-50 as image encoder. Parentheses show standard deviation across 3 random seeds. Missing values are denoted by –.
+
                             </p>
                         </div>
                     </div>
@@ -230,6 +231,10 @@ <h2 class="title is-3">Overall Results</h2>
                             <p>
                                 <b>Species Classification results on Snapshot Mountain Zebra dataset.</b>
                             </p>
+
+                            <br> <br>
+
+                            <b>Key Takeaways:</b> The addition of one or more contexts results in a performance gain over the no-context baseline in the vast majority of cases. Furthermore, the use of multiple contexts results in a performance boost in a majority of cases.
                         </div>
                     </div>
                 </div>
@@ -272,7 +277,7 @@ <h2 class="title is-3">Taxonomy-aware model results in more plausible prediction
                         <div class="item">
                             <img src="static/images/taxonomy_analysis.png" alt="Taxonomy-aware model results in more plausible predictions" />
                             <h2 class="subtitle">
-                                <b>(a):</b> Comparison of COSMO model with and without taxonomy edges. The use of taxonomy information helps the model to avoid semantically implausible predictions.
+                                <b>(a):</b> Comparison of COSMO model with and without taxonomy edges. The use of taxonomy information helps the model to avoid semantically implausible predictions.<br>
                                 <b>(b):</b> Quantitative evaluation of COSMO errors with and without taxonomy using a hierarchical distance metric (iWildCam2020-WILDS). The taxonomy-aware model achieves better performance in terms of Avg. LCA height.
                             </h2>