SGN-90206 International Doctoral Seminar: Signal Processing Journal Club
The Department of Signal Processing is organizing a Journal Club on Signal Processing
and related areas. The idea is to present
recent advances in signal processing and analysis, and most
presentations are based on an interesting research paper. The key idea
is to use minimum effort for
delivering maximum amount of
others. This is also an opportunity for the presenter to read a paper
that has always been interesting but there never was enough time to
read it. Alternatively, you can present your own work.
In order to follow the minimum effort principle, we prefer you not to
spend your time on making powerpoints. Instead, take the PDF and
highlight and comment the essential things with Adobe Acrobat tools or
Credit points can be gained from the seminar according to the following criteria:
- 1 presentation and 4 attendances gives one credit point.
- The seminar will continue next year, and you can increment the attendances until up to 5 credits.
- You can end the seminar and redeem the collected credits whenever you want.
What to present?
There will be a list of papers for presentation in the first seminar
meeting. You can also propose papers for presentation outside this list
(including your own work).
- Wednesday, February 11th at 13:15-14:00 (TB222): Ke Chen presents his paper Learning to Count
with Back-Propagated Information
- Monday, August 4th at 14:15 in TB222: Dr. Budhachandra Khundrakpam from Montreal Neurological
Institute, Montreal, Canada presents his work:
"Understanding Network Connectivity in the Developing Human Brain:
Implications for Cognitive Development and Neurodevelopmental
- Thursday, August 28th at 14:15 in TB224: Heikki Huttunen will
present the new software installed at Merope: pylearn2, scikit-learn, and OpenCV.
After that, we describe our 2nd place solution to the DecMeg2014
competition on Decoding the Human Brain. [PDF] [Code] [News]
- Tuesday, September 2nd at 14:15 in TB222: Giacomo Boracchi
(Politecnico di Milano, Italy): Change
and Anomaly Detection with Sparse Representations (ICASSP 2014) [abstract]
- Wednesday, 29.10 at 15:15 (TB222):
- Tuesday, 4.11 at 15:15 (TB224):
P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan
Object Detection with Discriminatively Trained Part Based Models
IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol.
32, No. 9, September 2010
- This work was awarded the PASCAL VOC "Lifetime Achievement" Prize in 2010.
- Tuesday, 11.11 at 15:15 (TB224):
Tuesday, 18.11 at 15:15 (TB224):
Thursday, 11.12 at 13:15 (TB224):
Prof.Slava Voloshinovsky will present his work. [abstract and biography]
6.2.2014 at 14:15-15:00 in
TB214 [Note changed time due to TUT ROCKS event]: Heikki Huttunen: Deep
learning for Signal Processing. Presentation of the paper:
- "Introduction to R"
Fri. 21.3.2014 at 14:15-15:00 in
TB220: Sakira Hassan presents the paper P. Geurts, D. Ernst., and L. Wehenkel, "Extremely randomized trees", Machine Learning, 63(1), 3-42, 2006.
The method is implemented as a part of the scikits-learn package.
Fri. 28.3.2014 at 14:15-15:00 in
Huttunen: Presentation of the use of Merope computing cluster at TUT.
Example: how to train a deep neural network on the Merope Tesla GPU
3.4.2014 at 15:15-16:00 in
TB220: student presentation
10.4.2014 at 15:15-16:00 in
TB220: Junsheng Fu presents the paper: Noah Snavely, Steven M. Seitz, Richard Szeliski. Photo Tourism:
Exploring image collections in 3D. ACM Transactions on Graphics
(Proceedings of SIGGRAPH 2006), 2006. See also supplementary website.
8.5.2014 at 15:15-16:00 in
TB220: Michal Joachimiak presents the paper: Hannuksela, M.M.;
Rusanovskyy, D.; Wenyi Su; Lulu Chen; Ri Li; Aflaki, P.; Deyan Lan;
Joachimiak, M.; Houqiang Li; Gabbouj, M., "Multiview-Video-Plus-Depth Coding
Based on the Advanced Video Coding Standard," IEEE Transactions on Image
Processing, Sept. 2013
22.4.2013 at 14:15-15:00 in TB223: Giacomo Boracchi
di Milano, Italy): Survey of Change Point Detection Methods.
21.3.2013 at 14:15-15:00 in TB216: Ana Sovic: "Fast Least
Absolute Deviation Adaptive Wavelet Filter Bank".
14.3.2013 at 13:15-14:00 in
TB224: Norbert Krueger from Cognitive
Vision Lab of Syddansk Universitet: Vision for Cognitive Robot
12.2.2013 at 12:15-13:00 in
TB222: Umar Iqbal and Srikanth Gopalakrishna: One shot
learning gesture recognition from RGB-D images. Their work is based
on this paper,
participated in the ChaLearn
gesture recognition challenge.
30.1.2013 at 14:15-15:00 in TB110:
Anssi Klapuri: Time-frequency and time-pitch representations for
4.12.2012 at 13:15-14:00 in TB224
Joni Kämäräinen: Gabor Features in Computer Vision and Image
20.11.2012 at 13:15-14:00 in
TB224 Anders Glent Buch: Object pose estimation for robot
23.8.2012 at 13:15-14:00 in TB223
- Jones, N., "Computer Science: The learning machines," Nature, Jan. 2014
- Hinton, Deng,
Yu, Dahl, Mohamed,
Jaitly, Senior, Vanhoucke, Nguyen, Sainath, Kingsbury, "Deep Neural
Networks for Acoustic Modeling in Speech Recognition," IEEE
Signal Processing Magazine, November 2012.
- Yoshua Bengio, "Deep Learning of Representations: Looking Forward," arXiv:1305.0445 [cs.LG]
24.10.2011 at 15:15-16:00 in
TB220 Juho Vihonen: R.
Mahony, T. Hamel, and J-M. Pflimlin, "Nonlinear complementary filters
on the special orthogonal group," IEEE Transactions on Automatic
Control, vol. 53, no. 5, June 2008.
7.11.2011 at 13:15-14:00 in TB215
"Logistic Regression for AML Prediction," best performing submission to
challenge. Tapio's submission is Team #347.
21.11.2011 at 15:15-16:00 in
TB220 Pekka Ruusuvuori: Analysis of brightfield focus image
5.12.2011 at 15:15-16:00 in
TB220 Presentation postponed by 2 weeks.
19.12.2011 at 15:15-16:00 in
TB220 Jukka-Pekka Kauppi: "Bilinear
Analysis" by Mads Dyrholm, Christoforos Christoforou, Lucas C.
Parra, Journal of Machine Learning Research 8 (2007) 1097-1111.
Energy Functions can be Minimized via Graph Cuts?
(Kolmogorov and Zabih, ECCV '02/PAMI
- Demonstration of the new Lytro
camera (15 min).
- Presentation of our recent submissions to the DREAM6
challenge (best performer) and the Amazon data science
challenge of IEEE MLSP 2012 (second place). Both are based on L1
regression classifier (30 min).
comprehensive panel of
three-dimensional models for studies of prostate cancer growth,
invasion and drug responses, (H\E4rm\E4, Happonen,
Kallioniemi et al. PLoS One vol. 5, 2010). Note: the place is
this time TC314.
(University of Jyv\E4skyl\E4, Department of Mathematics and Statistics): A
stochastic shape model for fibres with an
application to carbon nanotubes.
Niemist\F6: topic to be specified later. Cancelled.
Jussi Tohka: Presentation of Jussi's recent papers, e.g., J.V. Manjon,
J. Tohka , M. Robles. Improved estimates of partial volume coefficients
from noisy brain MRI using spatial context. NeuroSignal , in press,
Ruusuvuori: Worz, S.; Sander, P.; Pfannmoller, M.; Rieker, R.J.; Joos,
S.; Mechtersheimer, G.; Boukamp, P.; Lichter, P.; Rohr, K.; , "3D
Geometry-Based Quantification of Colocalizations in Multichannel 3D
Microscopy Signals of Human Soft Tissue Tumors," Medical Imaging,
IEEE Transactions on, vol.29, no.8, pp.1474-1484, Aug. 2010.
7.2.2011 at 16:15-17:00
Introduction to KFDA using the classic paper by Mika et al.: Fisher
discriminant analysis with kernels
(in IEEE NN for SP conf., 1999).
After that a presentation of: Heikki
Huttunen, Jari-Pekka Ryyn\E4nen, Heikki Forsvik, Ville Voipio and
Hisakazu Kikuchi, "Kernel Fisher Discriminant and Elliptic Shape Model
for Automatic Measurement of Allergic Reactions", Lecture Notes in
Computer Science, 2011, Volume 6688/2011, 764-773.
21.2.2011 at 15:15-16:00
High-throughput tracking of single yeast cells in a microfluidic
single-cell imaging platform Based on paper: D. Falconnet, A.
Niemist\F6, R. J. Taylor, M. Ricicova, T. Galitski, I. Shmulevich, and C.
L. Hansen, "High-throughput
tracking of single yeast cells in a microfluidic imaging matrix,"
Lab on a Chip, vol. 11, no. 3, pp. 466-473, 2011.
7.3.2011 at 15:15-16:00
Tomographic reconstruction from the incomplete projection data
21.3.2011 at 14:15-15:00: Dr. Izumi Ito from Tokyo
Technology will present the topic DCT sign correlation and its
application to signal matching.
4.4.2011 at 15:15-16:00 Heikki Huttunen: J.
Friedman, T. Hastie and R. Tibshirani, "Regularized Paths for
Generalized Linear Models via Coordinate Descent," Journal of
Statistical Software 33(1), 2010. After that,
a brief description of our
submission to ICANN MEG
classification competition, where we used the above technique
(a.k.a. logistic regression).
18.4.2011 at 15:15-16:00 Timo Erkkil\E4: Eugene
Tuv, Alexander Borisov, George Runger, Kari Torkkola, "Feature
with Ensembles, Artificial Variables, and Redundancy Elimination", Journal
of Machine Learning Research, Jan. 2009. Link to Timo's C++
Page created 14.2.07.
Last update 26.3.2014.