Händelser: Signaler och system, Elektroteknikhttp://www.chalmers.se/sv/om-chalmers/kalendariumAktuella händelser på Chalmers tekniska högskolaWed, 13 Jan 2021 09:06:45 +0100http://www.chalmers.se/sv/om-chalmers/kalendariumhttps://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/Masterpresentation-Gertz-Lindmark.aspxhttps://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/Masterpresentation-Gertz-Lindmark.aspxLouise Gertz, MPBME, och Albin Lindmark, MPSYS<p>Webseminarium</p><p>​ Knowledge distillation for face recognition on lightweight neural networks</p><div>​</div> <div><a href="https://chalmers.zoom.us/j/67565545895">Anslut till seminariet via Zoom.</a></div> <div>Lösenord: 214931</div> <div><br /></div> <div>Examinator: Marija Furdek, Inst för elektroteknik</div> <div>Handledare: Kenneth Jonsson, Smart Eye, Ahmet Oguz Kislal, Inst för elektroteknik</div> <div>Opponent: Jacob Larsson</div> <div><br /></div> <h2 class="chalmersElement-H2">Sammanfattning</h2> <div><br /></div> <div>Face recognition is a common bio-metric used in everyday commercial products and is also widely used in safety and surveillance. Accuracy is critical when face recognition is used for authentication purposes. Implementation of accurate face recognition using CNN models is limited to deployment in high-end complex systems due to the computational complexity. Viable implementation of accurate face recognition in mobile devices demands less computationally expensive methods, such as smaller models. This thesis investigates the potential of knowledge distillation (KD), a machine learning technique used to improve a small models performance by transferring knowledge from a large model to the smaller one. KD was implemented on CNNs trained for the task of face-identification and -verification using low resolution near infrared images. Both identification and verification models trained with KD achieved a higher accuracy than the reference models trained with standard procedures. Methods using a staged training procedure or hints comparing features of the models was shown to further improve KD and is useful when there is a large discrepancy between model sizes. Training using KD was proven to increase learning, thus making it possible to increase the accuracy of small face recognition networks.<br /></div>https://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/Carl-Toft,-Elektroteknik-1.aspxhttps://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/Carl-Toft,-Elektroteknik-1.aspxCarl Toft, Elektroteknik<p>online</p><p>​Titel: Towards Robust Visual Localization in Challenging Conditions</p><div><a href="https://chalmers.zoom.us/j/63019124068">​Anslut till disputationen från PC, Mac, Linux, iOS eller Android</a></div> <div><br /></div> <div>Maila till PhdAdm.e2@chalmers.se i god tid innan disputationen för att få lösenordet.</div> <br />The PhD defence can be accessed through Zoom, and the it will open shortly before 15:00. We would kindly ask you to keep the video off and mute the microphone during the seminar. At the end of the session there will be an opportunity to ask questions through Zoom. In case there will be any updates about the event, these will be posted on this website.<br /> ​<br />Carl Toft är doktorand vid forskargruppen Datorseende och medicinsk bildanalys<br />Opponent är Dr. Vladlen Koltun, Chief Scientist for Intelligent Systems, Intel, USA<br /><div>Examinator är Professor Fredrik Kahl vid forskargruppen <span>Datorseende och medicinsk bildanalys<span style="display:inline-block"></span></span></div> <div><br /></div> <div><br /></div> <div>Sammanfattning</div> <div>Visual localization is a fundamental problem in computer vision, with a multitude of applications in robotics, augmented reality and structure-from-motion. The basic problem is to, based on one or more images, figure out the position and orientation of the camera which captured these images relative to some model of the environment. Current visual localization approaches typically work well when the images to be localized are captured under similar conditions compared to those captured during mapping. However, when the environment exhibits large changes in visual appearance, due to e.g. variations in weather, seasons, day-night or viewpoint, the traditional pipelines break down. The reason is that the local image features used are based on low-level pixel-intensity information, which is not invariant to these transformations: when the environment changes, this will cause a different set of keypoints to be detected, and their descriptors will be different, making the long-term visual localization problem a challenging one.  <br /><br />In this thesis, five papers are included, which present work towards solving the problem of long-term visual localization. Two of the articles present ideas for how semantic information may be included to aid in the localization process: one approach relies only on the semantic information for visual localization, and the other shows how the semantics can be used to detect outlier feature correspondences. The third paper considers how the output from a monocular depth-estimation network can be utilized to extract features that are less sensitive to viewpoint changes. The fourth article is a benchmark paper, where we present three new benchmark datasets aimed at evaluating localization algorithms in the context of long-term visual localization. Lastly, the fifth article considers how to perform convolutions on spherical imagery, which in the future might be applied to learning local image features for the localization problem. <br /><br /></div> <br />https://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/Andreas-Buchberger,-Elektroteknik-2.aspxhttps://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/Andreas-Buchberger,-Elektroteknik-2.aspxAndreas Buchberger, Elektroteknik<p>online</p><p>​Titel: On Probabilistic Shaping and Learned Decoders with Application to Fiber-Optic Communications</p><div>​​<a href="https://chalmers.zoom.us/j/65353702126">Anslut till disputationen från PC, Mac, Linux, iOS eller Android<br /></a></div> <div>Maila till PhdAdm.e2@chalmers.se i god tid innan disputationen för att få lösenordet.<br /><br />The PhD defence can be accessed through Zoom, and the it will open shortly before 10:00. We would kindly ask you to keep the video off and mute the microphone during the seminar. At the end of the session there will be an opportunity to ask questions through Zoom. In case there will be any updates about the event, these will be posted on this website.<br /> ​<br />Andreas Buchberger är doktorand vid forskargruppen kommunikationssystem<br />Opponent är Professor Guido Montorsi, Politecnico di Torino, Italien<br />Examinator är Professor Alexandre Graell i Amat vid forskargruppen <span>kommunikationssystem<span style="display:inline-block"></span></span></div> <div><br /></div> <div>Sammanfattning</div> <div>Optical fibers form the backbone of today’s communication networks. Every phone call, text message, email, or website will, at some point, be transmitted through an optical fiber where light carries data from the transmitter to the receiver. Data is usually represented in binary form as a sequence of zeros and ones, both being equally likely (i.e., if we have a long sequence of bits, we know that about half of them are zero and half of them are one). However, it is known that transmitting such a stream of equiprobable bits is not optimal and does not result in the highest possible transmission rate. In the first part of this thesis, we design a system that transforms the sequence of equiprobable bits into a sequence where zeros are more likely than ones. This way, we can increase the transmission rate.</div> <div><br />Every communication system is subject to noise caused by different physical phenomena such as random thermal movements of electrons. When transmitting data, it hence may happen that we receive a zero when a one was sent and vice versa. By using error correcting codes, some of these transmission errors can be corrected at the receiver side. In the second part of this thesis, we use machine learning to implement these error correcting codes at the receiver and improve the reliability of the transmission.<br /><a href="https://chalmers.zoom.us/j/65353702126"></a></div>https://www.chalmers.se/sv/styrkeomraden/produktion/kalendarium/Sidor/Produktion-i-rymden.aspxhttps://www.chalmers.se/sv/styrkeomraden/produktion/kalendarium/Sidor/Produktion-i-rymden.aspxProduktion i rymden<p>RunAn, conference hall, Chalmersplatsen 1, Kårhuset</p><p> </p>​Den nya rymdåldern är här! Det tar vi fasta på i årets initiativseminarium som spinner vidare på 50-årsfirandet av månlandningen.​​<div><span style="background-color:initial">Hur tillverkar man för rymden? Vad forskar man på i rymden? Kan man tillverka produkter på Mars? Och hur bidrar Chalmers till rymdteknik och forskning? Det är några av de frågor vi kommer att lyfta under temat: Produktion i rymden.<br /></span> ​</div>https://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/International-Conference-on-Phantom-Limb-Pain-SV.aspxhttps://www.chalmers.se/sv/institutioner/e2/kalendarium/Sidor/International-Conference-on-Phantom-Limb-Pain-SV.aspxInternational Conference on Phantom Limb Pain<p>R-huset, Mölndals sjukhus, Mölndal</p><p>​Den första internationella konferensen om fantomsmärta, Phantom Limb Pain, (ICPLP) sammanför framstående experter inom forskning och behandling av fantomsmärtor i ett forum för öppen diskussion om teorier och rön.</p>​<div>​Läs mer <br /></div> <div><a href="/sv/institutioner/e2/nyheter/Sidor/Forsta-forskarkonferensen-om-fantomsmartornas-gata.aspx"><img class="ms-asset-icon ms-rtePosition-4" src="/_layouts/images/ichtm.gif" alt="" />​<span style="background-color:initial">Första forskarkonferensen om fantomsmärtornas gåta <br /></span></a></div> <div><span style="background-color:initial"><br /></span></div> <div><span style="background-color:initial">Mer information och anmälan till konferensen</span></div> <div><a href="http://www.bnl.chalmers.se/wordpress/index.php/icplp-2020/" target="_blank"><img class="ms-asset-icon ms-rtePosition-4" src="/_layouts/images/icgen.gif" alt="" />ICPLP:s webbplats​</a></div> <div><br /></div> <div>Konferensen var ursprungligen planerad att hållas den 2-4 september 2020 men har flyttats fram i tiden på grund av coronapandemin.</div> <div><br /></div>