Skip to content

Commit 3a9fed0

Browse files
author
Kartik Narayan
committed
update website
1 parent 905b5e3 commit 3a9fed0

File tree

1 file changed

+90
-92
lines changed

1 file changed

+90
-92
lines changed

‎docs/index.html

+90-92
Original file line numberDiff line numberDiff line change
@@ -85,25 +85,6 @@ <h1 class="title is-2 publication-title"> <i>FaceXFormer</i> : A Unified Transfo
8585

8686
<div class="column has-text-centered">
8787
<div class="publication-links">
88-
<!-- PDF Link. -->
89-
<!-- <span class="link-block">
90-
<a href=""
91-
class="external-link button is-normal is-rounded is-dark" target="_blank">
92-
<span class="icon">
93-
<i class="fas fa-file-pdf"></i>
94-
</span>
95-
<span>Paper</span>
96-
</a>
97-
</span> -->
98-
<!-- <span class="link-block">
99-
<a href=""
100-
class="external-link button is-normal is-rounded is-dark">
101-
<span class="icon">
102-
<i class="fas fa-file-pdf"></i>
103-
</span>
104-
<span>Supplementary material</span>
105-
</a>
106-
</span> -->
10788
<span class="link-block">
10889
<a href="https://arxiv.org/abs/2403.12960"
10990
class="external-link button is-normal is-rounded is-dark">
@@ -139,7 +120,7 @@ <h1 class="title is-2 publication-title"> <i>FaceXFormer</i> : A Unified Transfo
139120

140121
<h2 class="title is-3">Motivation & Contribution</h2>
141122
<img src="./static/images/intro.png" alt="" border="0" height="600" width="1500">
142-
<img src="./static/images/intro_table.png" alt="" style="border:0; height:500px; width:1500px;">
123+
<img src="./static/images/intro_table.png" alt="" style="border:0; height:1200px; width:1500px;">
143124
<div class="content has-text-justified">
144125
<p>Comparison with representative methods under different task settings.
145126
<i>FaceXformer</i>
@@ -158,7 +139,8 @@ <h2 class="title is-3">Motivation & Contribution</h2>
158139
</li>
159140
<li> <i>FaceXformer</i> is an end-to-end unified model capable of handling a comprehensive
160141
range of facial analysis tasks such as face parsing, landmark detection, head pose
161-
estimation, attributes prediction, and estimation of age, gender, race, epxression and face
142+
estimation, attributes prediction, and estimation of age, gender, race, epxression and
143+
face
162144
visibility.
163145
</li>
164146
<li> It leverages a transformer-based encoder-decoder architecture where
@@ -174,9 +156,6 @@ <h2 class="title is-3">Motivation & Contribution</h2>
174156
</div>
175157
</div>
176158
</div>
177-
<!--/ Abstract. -->
178-
179-
<!-- Paper video. -->
180159

181160
<section class="section">
182161
<div class="container is-max-desktop">
@@ -187,19 +166,21 @@ <h2 class="title is-3"><i>FaceXformer</i> Framework</h2>
187166
<div class="content has-text-justified">
188167
<h5 class="subtitle has-text-centered"></h5>
189168

190-
<img src="./static/images/main_archi.png" alt="" border=0 height=500 width=1500></img></
191-
<p>
192-
Overview of <i>FaceXformer</i> framework. It employs an encoder-decoder
193-
architecture, extracting multi-scale features from the input face image <b>I</b>, and
194-
fusing them into a unified representation <b>F</b> via MLP-Fusion. Task tokens <b>T</b>
195-
are processed alongside face representation <b>F</b> in the decoder, resulting in
196-
refined
197-
task-specific tokens <b><span style="position: relative; display: inline-block;">
198-
T
199-
<span
200-
style="position: absolute; top: -7px; left: 0.6px; right: 0; font-size: smaller;">^</span>
201-
</span></b>. These refined tokens are then used for
202-
task-specific predictions by passing through the unified head.
169+
<img src="./static/images/main_archi.png" alt="" border=0 height=500 width=1500></img>
170+
<p>
171+
Overview of <i>FaceXformer</i> framework. It employs an encoder-decoder
172+
architecture, extracting multi-scale features from the input face image <b>I</b>,
173+
and
174+
fusing them into a unified representation <b>F</b> via MLP-Fusion. Task tokens
175+
<b>T</b>
176+
are processed alongside face representation <b>F</b> in the decoder, resulting in
177+
refined
178+
task-specific tokens <b><span style="position: relative; display: inline-block;">
179+
T
180+
<span
181+
style="position: absolute; top: -7px; left: 0.6px; right: 0; font-size: smaller;">^</span>
182+
</span></b>. These refined tokens are then used for
183+
task-specific predictions by passing through the unified head.
203184
</p>
204185

205186

@@ -217,88 +198,105 @@ <h5 class="subtitle has-text-centered"></h5>
217198
<h2 class="title is-3">Quantitative Results</h2>
218199
<div class="content has-text-justified">
219200
<h5 class="subtitle has-text-centered"></h5>
220-
<img src="./static/images/parsing.png" alt="" border=0 height=500
221-
width=1500></img> </div>
201+
<img src="./static/images/parsing.png" alt="" border=0 height=500 width=1500></img>
222202
<p>
223203
Comparison with specialized models and existing multi-task networks on Face Parsing.
224204
</p>
225-
<img src="./static/images/hpe_lnd_attr" alt="" border=0 height=500
226-
width=1500></img> </div>
227-
<p>
228-
Comparison with specialized models and existing multi-task networks on Headpose Estimation, Landmarks Detection and Attributes Prediction.
229-
</p>
230-
<img src="./static/images/results_others" alt="" border=0 height=500
231-
width=1500></img> </div>
205+
</div>
206+
</div>
207+
</div>
208+
<div class="columns is-centered has-text-centered">
209+
<div class="column is-four-fifths">
210+
<div class="content has-text-justified">
211+
<h5 class="subtitle has-text-centered"></h5>
212+
<img src="./static/images/hpe_lnd_attr.png" alt="" border=0 height=500 width=1500></img>
232213
<p>
233-
Comparison with specialized models and existing multi-task networks on Facial Expression Recognition, Face Visibilty, and Age Estimation.
214+
Comparison with specialized models and existing multi-task networks on Headpose
215+
Estimation,
216+
Landmarks Detection and Attributes Prediction.
234217
</p>
235218
</div>
236219
</div>
237220
</div>
238-
</section>
239-
240-
<section class="section">
241-
<div class="container is-max-desktop">
242-
<!-- Abstract. -->
243221
<div class="columns is-centered has-text-centered">
244222
<div class="column is-four-fifths">
245-
<h2 class="title is-3">Qualitative Results</h2>
246223
<div class="content has-text-justified">
247224
<h5 class="subtitle has-text-centered"></h5>
248-
<img src="./static/images/qualitative.png" alt="" border=0 height=500
249-
width=1500></img><p>
250-
Qualitative results of <i>FaceXFormer</i>
225+
<img src="./static/images/results_others.png" alt="" border=0 height=500
226+
width=1500></img>
227+
<p>
228+
Comparison with specialized models and existing multi-task networks on Facial
229+
Expression
230+
Recognition, Face Visibilty, and Age Estimation.
251231
</p>
252232
</div>
253233
</div>
254234
</div>
255235
</div>
256-
</section>
236+
</div>
237+
</section>
238+
<section class="section">
239+
<div class="container is-max-desktop">
240+
<!-- Abstract. -->
241+
<div class="columns is-centered has-text-centered">
242+
<div class="column is-four-fifths">
243+
<h2 class="title is-3">Qualitative Results</h2>
244+
<div class="content has-text-justified">
245+
<h5 class="subtitle has-text-centered"></h5>
246+
<img src="./static/images/qualitative.png" alt="" border=0 height=500 width=1500></img>
247+
<p>
248+
Qualitative results of <i>FaceXFormer</i>
249+
</p>
250+
</div>
251+
</div>
252+
</div>
253+
</div>
254+
</section>
257255

258256

259-
<section class="section" id="BibTeX">
260-
<div class="container content is-max-desktop">
261-
<h2 class="title">BibTeX</h2>
262-
<pre><code>@article{narayan2024facexformer,
257+
<section class="section" id="BibTeX">
258+
<div class="container content is-max-desktop">
259+
<h2 class="title">BibTeX</h2>
260+
<pre><code>@article{narayan2024facexformer,
263261
title={FaceXFormer: A Unified Transformer for Facial Analysis},
264262
author={Narayan, Kartik and VS, Vibashan and Chellappa, Rama and Patel, Vishal M},
265263
journal={arXiv preprint arXiv:2403.12960},
266264
year={2024}
267265
}
268266
</code></pre>
269-
</div>
270-
</section>
267+
</div>
268+
</section>
271269

272-
<section class="section">
273-
<div class="container is-max-desktop content">
274-
<h5 class="title" style="font-size: 10px;"> Acknowledgement: The website template is taken from
275-
<span class="author-block">
276-
<a href="https://nerfies.github.io/" target="_blank">Nerfies</a>
277-
</h5>
270+
<section class="section">
271+
<div class="container is-max-desktop content">
272+
<h5 class="title" style="font-size: 10px;"> Acknowledgement: The website template is taken from
273+
<span class="author-block">
274+
<a href="https://nerfies.github.io/" target="_blank">Nerfies</a>
275+
</h5>
278276

279-
</div>
280-
</section>
277+
</div>
278+
</section>
281279

282-
<script>
283-
const viewers = document.querySelectorAll(".image-compare");
284-
viewers.forEach((element) => {
285-
let view = new ImageCompare(element, {
286-
hoverStart: true,
287-
addCircle: true
288-
}).mount();
289-
});
290-
291-
$(document).ready(function () {
292-
var editor = CodeMirror.fromTextArea(document.getElementById("bibtex"), {
293-
lineNumbers: false,
294-
lineWrapping: true,
295-
readOnly: true
296-
});
297-
$(function () {
298-
$('[data-toggle="tooltip"]').tooltip()
299-
})
300-
});
301-
</script>
280+
<script>
281+
const viewers = document.querySelectorAll(".image-compare");
282+
viewers.forEach((element) => {
283+
let view = new ImageCompare(element, {
284+
hoverStart: true,
285+
addCircle: true
286+
}).mount();
287+
});
288+
289+
$(document).ready(function () {
290+
var editor = CodeMirror.fromTextArea(document.getElementById("bibtex"), {
291+
lineNumbers: false,
292+
lineWrapping: true,
293+
readOnly: true
294+
});
295+
$(function () {
296+
$('[data-toggle="tooltip"]').tooltip()
297+
})
298+
});
299+
</script>
302300
</body>
303301

304-
</html>
302+
</html>

0 commit comments

Comments
 (0)