adaptation of the web site

ENSTA-U2IS-AI · Jun 2, 2024 · 5a6e50b · 5a6e50b
1 parent 45d5337
commit 5a6e50b
Show file tree

Hide file tree

Showing 4 changed files with 44 additions and 58 deletions.
diff --git a/index.html b/index.html
@@ -118,9 +118,9 @@
 
 <html>
 <head>
-	<title>This is my paper title</title>
+	<title>Make Me a BNN</title>
 	<meta property="og:image" content="Path to my teaser.png"/> <!-- Facebook automatically scrapes this. Go to https://developers.facebook.com/tools/debug/ if you update and want to force Facebook to rescrape. -->
-	<meta property="og:title" content="Creative and Descriptive Paper Title." />
+	<meta property="og:title" content="Make Me a BNN: A Simple Strategy for Estimating Bayesian Uncertainty from Pre-trained Models." />
 	<meta property="og:description" content="Paper description." />
 
 	<!-- Get from Google Analytics -->
@@ -138,37 +138,54 @@
 <body>
 	<br>
 	<center>
-		<span style="font-size:36px">Creative and Descriptive Paper Title</span>
+		<span style="font-size:36px">Make Me a BNN: A Simple Strategy for Estimating
+Bayesian Uncertainty from Pre-trained Models</span>
 		<table align=center width=600px>
 			<table align=center width=600px>
 				<tr>
 					<td align=center width=100px>
 						<center>
-							<span style="font-size:24px"><a href="https://en.wikipedia.org/wiki/James_J._Gibson">First Author</a></span>
+							<span style="font-size:24px"><a href="https://giannifranchi.github.io/">Gianni Franchi</a></span>
 						</center>
 					</td>
 					<td align=center width=100px>
 						<center>
-							<span style="font-size:24px"><a href="https://en.wikipedia.org/wiki/James_J._Gibson">Second Author</a></span>
+							<span style="font-size:24px"><a href="https://scholar.google.fr/citations?user=RW4CQ68AAAAJ&hl=fr">Olivier Laurent</a></span>
 						</center>
 					</td>
 					<td align=center width=100px>
 						<center>
-							<span style="font-size:24px"><a href="https://en.wikipedia.org/wiki/James_J._Gibson">Third Author</a></span>
+							<span style="font-size:24px"><a href="">Maxence Leguery</a></span>
 						</center>
 					</td>
+					<td align=center width=100px>
+						<center>
+							<span style="font-size:24px"><a href="https://abursuc.github.io/">Andrei Bursuc</a></span>
+						</center>
+					</td>
+					<td align=center width=100px>
+						<center>
+							<span style="font-size:24px"><a href="https://scholar.google.it/citations?user=zooORRsAAAAJ&hl=it">Andrea Pilzer</a></span>
+						</center>
+					</td>
+					<td align=center width=100px>
+						<center>
+							<span style="font-size:24px"><a href="https://www.comp.nus.edu.sg/~ayao/">Angela Yao</a></span>
+						</center>
+					</td>
+
 				</tr>
 			</table>
 			<table align=center width=250px>
 				<tr>
 					<td align=center width=120px>
 						<center>
-							<span style="font-size:24px"><a href=''>[Paper]</a></span>
+							<span style="font-size:24px"><a href='https://arxiv.org/abs/2312.15297'>[Paper]</a></span>
 						</center>
 					</td>
 					<td align=center width=120px>
 						<center>
-							<span style="font-size:24px"><a href='https://github.com/richzhang/webpage-template'>[GitHub]</a></span><br>
+							<span style="font-size:24px"><a href='https://torch-uncertainty.github.io/'>[GitHub]</a></span><br>
 						</center>
 					</td>
 				</tr>
@@ -181,15 +198,15 @@
 			<tr>
 				<td width=260px>
 					<center>
-						<img class="round" style="width:500px" src="./resources/teaser.png"/>
+						<img class="round" style="width:500px" src="./resources/process_abnn.png"/>
 					</center>
 				</td>
 			</tr>
 		</table>
 		<table align=center width=850px>
 			<tr>
 				<td>
-					This was a template originally made for <a href="http://richzhang.github.io/colorization/">Colorful Image Colorization</a>. The code can be found in this <a href="https://github.com/richzhang/webpage-template">repository</a>.
+					Illustration of the training process for the ABNN. The procedure begins with training a single DNN omega (w) MAP ,followed by architectural adjustments to transform it into an ABNN. The final step involves fine-tuning the ABNN model.
 				</td>
 			</tr>
 		</table>
@@ -201,7 +218,9 @@
 		<center><h1>Abstract</h1></center>
 		<tr>
 			<td>
-				This is my abstract.
+				Deep Neural Networks (DNNs) are powerful tools for various computer vision tasks, yet they often struggle with reliable uncertainty quantification — a critical requirement for real-world applications. Bayesian Neural Networks (BNN) are equipped for uncertainty estimation but cannot scale to large DNNs where they are highly unstable to train.
+To address this challenge, we introduce the Adaptable Bayesian Neural Network (ABNN), a simple and scalable strategy to seamlessly transform DNNs into BNNs in a post-hoc manner with minimal computational and training overheads.
+ABNNpreserves the main predictive properties of DNNs while enhancing their uncertainty quantification abilities through simple BNN adaptation layers (attached to normalization layers) and a few fine-tuning steps on pre-trained models. We conduct extensive experiments across multiple datasets for image classification and semantic segmentation tasks, and our results demonstrate that ABNN achieves state-of-the-art performance without the computational budget typically associated with ensemble methods.
 			</td>
 		</tr>
 	</table>
@@ -210,65 +229,32 @@
 	<hr>
 	<center><h1>Talk</h1></center>
 	<p align="center">
-		<iframe width="660" height="395" src="https://www.youtube.com/embed/dQw4w9WgXcQ" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen align="center"></iframe>
+		<iframe id="player" type="text/html" width="640" height="360"
+  src="http://www.youtube.com/embed/aXqVBAOXc0o?enablejsapi=1"
+  frameborder="0"></iframe>
 	</p>
 
 	<table align=center width=800px>
 		<br>
 		<tr>
 			<center>
-				<span style="font-size:28px"><a href=''>[Slides]</a>
+				<span style="font-size:28px"><a href="./resources/CVPR-presentation_ABNN.pdf"/>[Slides]</a>
 				</span>
 			</center>
 		</tr>
 	</table>
 	<hr>
 
-	<center><h1>Code</h1></center>
-
-	<table align=center width=420px>
-		<center>
-			<tr>
-				<td>
-				</td>
-			</tr>
-		</center>
-	</table>
-	<table align=center width=400px>
-		<tr>
-			<td align=center width=400px>
-				<center>
-					<td><img class="round" style="width:450px" src="./resources/method_diagram.png"/></td>
-				</center>
-			</td>
-		</tr>
-	</table>
-	<table align=center width=850px>
-		<center>
-			<tr>
-				<td>
-					Short description if wanted
-				</td>
-			</tr>
-		</center>
-	</table>
-	<table align=center width=800px>
-		<br>
-		<tr><center>
-			<span style="font-size:28px">&nbsp;<a href='https://github.com/richzhang/webpage-template'>[GitHub]</a>
-			</center>
-		</span>
-	</table>
 	<br>
 	<hr>
 	<table align=center width=450px>
 		<center><h1>Paper and Supplementary Material</h1></center>
 		<tr>
 			<td><a href=""><img class="layered-paper-big" style="height:175px" src="./resources/paper.png"/></a></td>
-			<td><span style="font-size:14pt">F. Author, S. Author, T. Author.<br>
-				<b>Creative and Descriptive Paper Title.</b><br>
-				In Conference, 20XX.<br>
-				(hosted on <a href="">ArXiv</a>)<br>
+			<td><span style="font-size:14pt">G. Franchi, O. Laurent, M. Leguery, A. Bursuc, A. Pilzer, A. Yao<br>
+				<b>Make Me a BNN: A Simple Strategy for Estimating Bayesian Uncertainty from Pre-trained Models.</b><br>
+				In IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR), 2024.<br>
+				(hosted on <a href="https://arxiv.org/html/2312.15297v1">ArXiv</a>)<br>
 				<!-- (<a href="./resources/camera-ready.pdf">camera ready</a>)<br> -->
 				<span style="font-size:4pt"><a href=""><br></a>
 				</span>

diff --git a/resources/CVPR-presentation_ABNN.pdf b/resources/CVPR-presentation_ABNN.pdf
diff --git a/resources/bibtex.txt b/resources/bibtex.txt
@@ -1,6 +1,6 @@
-@inproceedings{author20XXtitle,
-  title={Please cite me},
-  author={Author, First and Author, Second and Author, Third},
-  booktitle={Conference},
-  year={20XX}
-}
+@article{franchi2023make,
+  title={Make Me a BNN: A Simple Strategy for Estimating Bayesian Uncertainty from Pre-trained Models},
+  author={Franchi, Gianni and Laurent, Olivier and Legu{\'e}ry, Maxence and Bursuc, Andrei and Pilzer, Andrea and Yao, Angela},
+  journal={arXiv preprint arXiv:2312.15297},
+  year={2023}
+}
diff --git a/resources/process_abnn.png b/resources/process_abnn.png