Skip to content

Commit 75e0c7e

Browse files
amonszpartamonszpart
committed
[map-free-workshop] Torsten talk summary
1 parent 8e186e9 commit 75e0c7e

File tree

2 files changed

+99
-17
lines changed

2 files changed

+99
-17
lines changed

‎map-free-workshop/2024/css/style.css

Lines changed: 52 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -138,4 +138,55 @@ hr {
138138

139139
.workshop-paper {}
140140
.extended-abstract {}
141-
.method-description {}
141+
.method-description {}
142+
143+
.iframe-frame-radial {
144+
position: absolute;
145+
top: -40px;
146+
left: -40px;
147+
/*right: 40px;*/
148+
/*bottom: 40px;*/
149+
background: radial-gradient(
150+
ellipse at center,
151+
rgba(0, 0, 0, 0) 0%,
152+
rgba(0, 0, 0, 0) calc(100% - 49px),
153+
rgba(0, 0, 0, 1) calc(100% - 50px),
154+
white 100%);
155+
background-size: 100% 100%, 100% 100%;
156+
background-repeat: no-repeat;
157+
background-position: center;
158+
width: calc(100% + 80px);
159+
height: calc(100% + 80px);
160+
}
161+
162+
.iframe-frame-tb {
163+
position: absolute;
164+
top: 20px;
165+
left: 20px;
166+
background: linear-gradient(to bottom, white, black 20px, rgba(0, 0, 0, 0.0) 21px, rgba(0, 0, 0, 0.0) calc(100% - 21px), black calc(100% - 20px), white);
167+
background-size: 100% 100%, 100% 100%;
168+
background-repeat: no-repeat;
169+
background-position: center;
170+
width: calc(100% - 40px);
171+
height: calc(100% - 40px);
172+
}
173+
174+
.iframe-frame-lr {
175+
position: absolute;
176+
top: 40px;
177+
left: -20px;
178+
background: linear-gradient(to right, white, black 20px, rgba(0, 0, 0, 0.0) 21px, rgba(0, 0, 0, 0.0) calc(100% - 21px), black calc(100% - 20px), white);
179+
background-size: 100% 100%, 100% 100%;
180+
background-repeat: no-repeat;
181+
background-position: center;
182+
height: calc(100% - 80px);
183+
width: calc(100% + 40px);
184+
}
185+
.yourClass {
186+
display: inline-flex;
187+
border: double 10px transparent;
188+
border-radius: 20px;
189+
background-image: radial-gradient(circle at center, #fff 0%, #fff 90%, #000 100%);
190+
background-origin: border-box;
191+
background-clip: content-box, border-box;
192+
}

‎map-free-workshop/2024/index.html

Lines changed: 47 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -207,10 +207,10 @@ <h4 class="speaker-name">
207207
--></a>, Meta
208208
</h4>
209209
<p class="speaker-bio">
210-
<span class="talk-title">Talk title:</span> TBC
210+
<span class="talk-title">Talk title</span>: TBC
211211
</p>
212212
<p class="speaker-bio">
213-
<span class="talk-title">Bio:</span> Jakob Engel joined the Surreal Vision team at Oculus
213+
<span class="talk-title">Bio</span>: Jakob Engel joined the Surreal Vision team at Oculus
214214
Research in Redmond in 2016, working on the future of 3D-enabled Machine Perception. He did his
215215
Bachelor and Master at TU Munich (2009 and 2012), followed up with a PhD at the Computer Vision
216216
Group there, headed by Professor Daniel Cremers. He spent 6 months as Intern at Intel Research
@@ -231,16 +231,16 @@ <h4 class="speaker-name">
231231
--></a>, Naver Labs Europe
232232
</h4>
233233
<p class="speaker-bio">
234-
<span class="talk-title">Talk title:</span> Grounding Image Matching in 3D with MASt3R <br/>
235-
<span class="talk-title">Summary:</span> The journey from CroCo to MASt3R exemplify a
234+
<span class="talk-title">Talk title</span>: Grounding Image Matching in 3D with MASt3R <br/>
235+
<span class="talk-title">Summary</span>: The journey from CroCo to MASt3R exemplify a
236236
significant paradigm shift in 3D vision technologies. This presentation will delve into the
237237
methodologies, innovations, and synergistic integration of these frameworks, demonstrating their
238238
impact on the field and potential future directions. The discussion aims to highlight how these
239239
advancements unify and streamline the processing of 3D visual data, offering new perspectives
240240
and capabilities in map-free visual relocalization, robotic navigation and beyond.
241241
</p>
242242
<p class="speaker-bio">
243-
<span class="talk-title">Bio:</span> Vincent is a research scientist in Geometric Deep Learning
243+
<span class="talk-title">Bio</span>: Vincent is a research scientist in Geometric Deep Learning
244244
at Naver Labs Europe.
245245
He joined 5 years ago, in 2019, after completing his PhD on Multi-View Stereo Reconstruction for
246246
dynamic shapes at the INRIA Grenoble-Alpes under the supervision of E. Boyer and J-S. Franco.
@@ -261,10 +261,10 @@ <h4 class="speaker-name">
261261
--></a>, Google
262262
</h4>
263263
<p class="speaker-bio">
264-
<span class="talk-title">Talk title:</span> TBC
264+
<span class="talk-title">Talk title</span>: TBC
265265
</p>
266266
<p class="speaker-bio">
267-
<span class="talk-title">Bio:</span> Simon Lynen is a tech lead manager at Google Zurich. His
267+
<span class="talk-title">Bio</span>: Simon Lynen is a tech lead manager at Google Zurich. His
268268
group focuses on providing high precision mobile-phone localization as part of the Visual
269269
Positioning Service (VPS). Devices with Google’s augmented reality capabilities can leverage VPS
270270
to enable global scale location aware experiences such as ARCore CloudAnchors and GoogleMaps
@@ -287,10 +287,16 @@ <h4 class="speaker-name">
287287
--></a>, CTU Prague
288288
</h4>
289289
<p class="speaker-bio">
290-
<span class="talk-title">Talk title:</span> TBC
290+
<span class="talk-title">Talk title</span>: Scene Representations for Visual Localization <br/>
291+
<span class="talk-title">Summary</span>: Visual localization is the problem of estimating the
292+
exact position and orientation from which a given image was taken. Traditionally, localization
293+
approaches either used a set of images with known camera poses or a sparse point cloud, obtained
294+
from Structure-from-Motion, to represent the scene. In recent years, the list of available scene
295+
representations has grown considerably. In this talk, we review a subset of the available
296+
representations.
291297
</p>
292298
<p class="speaker-bio">
293-
<span class="talk-title">Bio:</span> Torsten Sattler is a Senior Researcher at CTU. Before, he
299+
<span class="talk-title">Bio</span>: Torsten Sattler is a Senior Researcher at CTU. Before, he
294300
was a tenured associate professor at Chalmers University of Technology. He received a PhD in
295301
Computer Science from RWTH Aachen University, Germany, in 2014. From Dec. 2013 to Dec. 2018, he
296302
was a post-doctoral and senior researcher at ETH Zurich. Torsten has worked on feature-based
@@ -316,9 +322,9 @@ <h4 class="speaker-name">
316322
--></a>, Carnegie Mellon University
317323
</h4>
318324
<p class="speaker-bio">
319-
<span class="talk-title">Talk title:</span> Rethinking Camera Parametrization for Pose
325+
<span class="talk-title">Talk title</span>: Rethinking Camera Parametrization for Pose
320326
Prediction<br/>
321-
<span class="talk-title">Summary:</span> Every student of projective geometry is taught to
327+
<span class="talk-title">Summary</span>: Every student of projective geometry is taught to
322328
represent camera matrices via an extrinsic and intrinsic matrix and learning-based methods that
323329
seek to predict viewpoints given a set of images typically adopt this (global) representation.
324330
In this talk, I will advocate for an over-parametrized local representation which represents
@@ -327,7 +333,7 @@ <h4 class="speaker-name">
327333
for neural learning and lead to more accurate camera prediction.
328334
</p>
329335
<p class="speaker-bio">
330-
<span class="talk-title">Bio:</span> Shubham Tulsiani is an Assistant Professor at Carnegie
336+
<span class="talk-title">Bio</span>: Shubham Tulsiani is an Assistant Professor at Carnegie
331337
Mellon University in the Robotics Institute. Prior to this, he was a research scientist at
332338
Facebook AI Research (FAIR). He received a PhD. in Computer Science from UC Berkeley in 2018
333339
where his work was supported by the Berkeley Fellowship. He is interested in building perception
@@ -347,10 +353,10 @@ <h4 class="speaker-name">
347353
--></a>, Niantic and Oxford University
348354
</h4>
349355
<p class="speaker-bio">
350-
<span class="talk-title">Talk title:</span> TBC (opening remarks)
356+
<span class="talk-title">Talk title</span>: TBC (opening remarks)
351357
</p>
352358
<p class="speaker-bio">
353-
<span class="talk-title">Bio:</span> Professor Victor Adrian Prisacariu received the Graduate
359+
<span class="talk-title">Bio</span>: Professor Victor Adrian Prisacariu received the Graduate
354360
degree (with first class hons.) in computer engineering from Gheorghe Asachi Technical
355361
University, Iasi, Romania, in 2008, and the D.Phil. degree in engineering science from the
356362
University of Oxford in 2012.<br/>
@@ -464,7 +470,14 @@ <h4>📅&nbsp; &nbsp; &nbsp; On Monday, 30<sup>th</sup> September, 2024, AM </h4
464470
</tr>
465471
<tr>
466472
<td class="text-nowrap">11:50 - 12:20</td>
467-
<td><a href="https://tsattler.github.io/">Torsten Sattler</a>, CTU Prague</td>
473+
<td><a href="https://tsattler.github.io/">Torsten Sattler</a>, CTU Prague<br/>
474+
<i>Scene Representations for Visual Localization</i><br/>
475+
Visual localization is the problem of estimating the exact position and orientation from which a
476+
given image was taken. Traditionally, localization approaches either used a set of images with
477+
known camera poses or a sparse point cloud, obtained from Structure-from-Motion, to represent
478+
the scene. In recent years, the list of available scene representations has grown considerably.
479+
In this talk, we review a subset of the available representations.
480+
</td>
468481
</tr>
469482
<tr>
470483
<td class="text-nowrap">12:25 - 12:55</td>
@@ -481,7 +494,25 @@ <h4>📅&nbsp; &nbsp; &nbsp; On Monday, 30<sup>th</sup> September, 2024, AM </h4
481494
</tr>
482495
<tr>
483496
<td class="text-nowrap">12:55 - 13:00</td>
484-
<td>Closing Remarks and award photos</td>
497+
<td>
498+
<div class="table-responsive">
499+
<table class="table table-borderless p-0 m-0">
500+
<tr class="p-0 m-0">
501+
<td class="p-0 m-0 col-4">Closing Remarks and award photos</td>
502+
<td class="p-0 m-0 col-1"></td>
503+
<td class="p-0 m-0 col-6">
504+
<div class="ratio ratio-1x1 text-center align-middle">
505+
<iframe class="yourClass"
506+
style="border: 0; width: calc(100% - 0px); height: calc(100% - 0px);"
507+
allowfullscreen
508+
src="https://scaniverse.com/scan/qafjd2g35pd7awe4?embed=1"></iframe>
509+
</div> <!-- ratio -->
510+
</td>
511+
<td class="p-0 m-0 col-1"></td>
512+
</tr>
513+
</table>
514+
</div> <!-- table-responsive -->
515+
</td>
485516
</tr>
486517
</tbody>
487518
</table>

0 commit comments

Comments
 (0)