Second semester of 2016

Class hours: mondays and wednesdays from 5:30pm to 7:30pm

Room: PC-05 (theoretical classes) and Lab. 3 (practical classes)

email list: https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/ci396-menotti

Once you subscribe to the list a message asking your confirmation is sent. Only after this confirmation you will effectively be subscribed.

The list accepts only messages sent from the email address that you are subscribed.

More than one subscription is allowed (with different email address).

Evaluation

Two written tests:
	40 points - 20 points each
Ten practical tests:
	40 points - 4 points each
A seminar/workshop:
	20 points - slides + presentation (30/45 minutes depending on the number of attending students)

Classical Problems

Face Recognition
Iris Recognition
Fingerprint Recognition
Pedestrian Detection
Detection -- Face (License plate, Vehicle, Eye & Iris)
Iris Segmentation
Face Fiducial Points Detection
Person Action Recognition
First-Person Activity Recognition
Spoofing Detection

Proposed Schedule

##

day

Content

01

08/01

Introduction of the course

02

08/03

Problem 01: Face Recognition (Theoretical) [slides1/slides2]

03

08/08

Problem 01: Face Recognition (Practical) [activity1] - Loading databases & Meanfaces

04

08/10

Problem 01: Face Recognition (Practical) [activity1] - Eigenfaces

05

08/15

Problem 01: Face Recognition (Practical) [activity1] - Classification

06

08/17

Problem 02: Face Recognition (Practical) [activity1] - Evaluation/validation

07

08/22

Problem 02: Iris Recognition (Theoretical) [slides]

08

08/24

Problem 02: Iris Recognition (Practical) [activity2] - Loading databases & extracting pupil region

09

08/29

Problem 02: Iris Recognition (Practical) [activity2] - Extracting iris region and normalization

10

08/31

Problem 02: Iris Recognition (Practical) [activity2] - Computing features (Wavelet & LBP) for Iris Verification & Identification

11

09/05

Problem 02: Iris Recognition (Practical) [activity2] - Computing metrics (FAR/FRR/EER/ROC Curve & Accuracy) for Iris Verification & Identification

--

09/07

No classes -- Holyday: Dia da Independência (Independence Day)

12

09/12

1st written test

13

09/14

Problem 03: Fingerprint Recognition (Theoretical) [slides]

14

09/19

Problem 03: Fingerprint Recognition (Practical) - Loading databases & Image enhancement

15

09/21

Problem 03: Fingerprint Recognition (Practical) - Orientation Maps & Type Annotation

16

09/26

Problem 03: Fingerprint Recognition (Practical) - Region of Interest Detection & Singular Point Detection

17

09/28

Problem 03: Fingerprint Recognition (Practical) - Singular Point Detection & Type Classification

18

10/03

Problem 03: Fingerprint Recognition (Practical) - Thinning & Minutiae Detection

--

10/05

No classes -- Semana Integrada de Ensino, Pesquisa e Extensão (SIEPE)

19

10/10

Problem 03: Fingerprint Recognition (Practical) - Minutiae Extraction

--

10/12

No classes -- Holyday: Padroeira do Brasil (Patroness of Brazil)

20

10/17

Problem 03: Fingerprint Recognition (Practical) - Pattern Matching / Score Computation

21

10/19

Problem 04: Pedestrian Detection (Theorectical) [slides1/slides2/slides3]

--

10/24

No classes -- Undergraduate Computer Science Week

--

10/26

No classes -- Undergraduate Computer Science Week

22

10/31

Problem 04: Pedestrian Detection (Practical) - Loading databases & Image Pyramids

--

11/02

No classes -- Holyday: Finados (All souls)

23

11/07

Problem 04: Pedestrian Detection (Practical) - HOG computation

24

11/09

Problem 04: Pedestrian Detection (Practical) - Training SVM

25

11/14

Problem 04: Pedestrian Detection (Practical) - Hard negative mining

26

11/16

Problem 04: Pedestrian Detection (Practical) - Sliding windows & Non-maximum suppression

27

11/21

Problem 04: Pedestrian Detection (Practical) - curve miss rate vs FPPW

28

11/23

2nd written tests

29

11/28

Presentation of grades for written tests and practical activities

29

11/28

Seminars:	Multimodality - face+iris - Luiz Zanlorensi [slides] Thermal images as source for biometrics - Eric G. Müller [slides]

30

11/30

Seminars:	Multimodality - iris+eye - Cides Bezerra [slides] Palm print - Pedro M. S. Pereira [slides]

31

12/05

Seminars:	EEG brainwaves - Jeovane H. Alves [slides] Hand vascular vein - Anderson T. S. Igarashi [slides] First-person activity recognition - Matheus A. Nerone [slides]

32

12/07

Seminars:	Person re-identification - Guilherme Z. Machado [slides]

33

12/12

Seminars:	Object tracking - Renan L. Burda [slides] Person action recognition - Fabricio J. O. Ceschin [slides]

--

12/14

Preparation for final exam

--

12/19

Final exam

24

01/09

(strike) Problem 04: Pedestrian Detection (Practical) - Training SVM

25

01/11

(strike) Problem 04: Pedestrian Detection (Practical) - Hard negative mining

26

01/16

(strike) Problem 04: Pedestrian Detection (Practical) - Sliding windows & Non-maximum suppression

27

01/18

(strike) Seminars:
	Ear as biometrics - Clarissa D. Pereira [slides] Lip recognition - Victor Mocelin [slides]

28

01/23

(strike) Problem 04: Pedestrian Detection (Practical) - curve miss rate vs FPPW

29

01/25

(strike) Practical Activity #4 -- submission & evaluation

30

01/30

(strike) 2nd written tests (already done!)

--

02/06

(strike) Final exam

Support material

Nothing <???.pdf>

References

Face Verification
- Eigenfaces for Recognition, M. Turk and A.Pentland, Journal of cognitive Neuroscience, 1991, vol. 3, no. 1, pages 71-86 -- 14,985 citations on scholar.google.com on 2016/08/03.
- Eigenfaces vs. Fisherfaces: recognition using class specific linear projection, P. N. Belhumeur, J. P. Hespanha and D. J. Kriegman, IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997, vol. 19, no. 7, pages 711-720 -- 10,909 citations on scholar.google.com on 2016/08/03.
Iris Recognition
- How Iris Recognition Works, J. Daugman, IEEE Transactions on Circuits and Systems for Video Technology, 2004, vol. 14, no. 1, pages 21-30 -- 3,242 citations on scholar.google.com on 2016/08/22.
- Iris Recognition: An Emerging Biometric Technology, R. P. Wildes, IEEE Proceedings of IEEE, 1997, vol. 85, no. 9, pages 1348-1363 -- 2,334 citations on scholar.google.com on 2016/08/22
- Toward Noncooperative Iris Recognition: A classification Approach Using Multiple Signatures, H. Proenca and L. A. Alexandre, IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, vol. 29, no. 4, pages 607-612 -- 191 citations on scholar.google.com on 2016/08/22
Fingerprint Recognition
- On-line fingerprint verification, A. Jain, L. Hong, R. Bolle, IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997, vol. 19, no. 4, pages 302-314 -- 1,778 citations on scholar.google.com on 2016/09/19
- Fingerprint recognition using wavelet features, M. Tico, E. Immonen, P. Ramo, P. Kuosmanen and J. Saarinen, IEEE International Symposium on Circuits and Systems, 2001, vol. 2, pages 21-24 -- 98 citations on scholar.google.com on 2016/09/19
- Implementation of an automatic fingerprint identification system, P. Huang, C.-Y. Chang and C.-C. Chen, IEEE International Conference on Electro/Information Technology, 2007, pages 461-464 -- 4 citations on scholar.google.com on 2016/09/19
- Classificaiton of Fingerprint Images, L. Hong and A. Jain, Scandinaval Conference on Image Analysis, 1999, vol. 2, pages 1-7 -- 164 citations on scholar.google.com on 2016/09/19
Pedestrian Detection
- Histograms of Oriented Gradients for Human Detection, N. Dalal and B. Triggs, IEEE Conference on Computer Vision and Pattern Recognition, 2005, pages 1-8 -- 15,999 citations on scholar.google.com on 2016/10/17
- Finding People in Images and Videos, N. Dalal, PhD Thesis, INRIA Rhône-Alpes / Institut National Polytechnique de Grenoble, 2006, 135 pages -- 353 citations on scholar.google.com on 2016/10/19
- Fast Human Detection Using a Cascade of Histograms of Oriented Gradients, Q. Zhu, S. Avidan, M-C. Yeh and K-T. Cheng, IEEE Conference on Computer Vision and Pattern Recognition, 2006, 1-8, -- 1,365 citations on scholar.google.com on 2016/10/17

Databases

Face Verification
- Harvard Robotics Laboratory --
- ORL (AT&T) -- 10 subjects x 40 images, each one with 92x112 pixels and 256 grey levels
- The Yale Face Database -- 15 subjects x 11 images, each one with 320 x 243 pixels
- The Yale Face Database B -- 10 subjects x 576 viewing conditions (9 poses x 64 illuminations) -- 5760+90=5850 images
- The Extended Yale Face Database B -- 28 humans under 9 poses and 64 illumination conditions (adding 18 subjects)
- PubFig83 -- 83 subjects x 100 images, each with 100x100 pixels
Face Expression Recognition
- The Yale Face Database -- 15 subjects x 11 images, each one with 320 x 243 pixels
Iris Recognition
- Chinese Academy of Sciences Institue of Automation (CASIA) --
  
  CASIA-IrisV4 Interval
  
  CASIA-IrisV4 Lamp
- University Beira Interior -- Noisy Visible Wavelength Iris (UBIRIS)
  
  UBIRIS.v2
Fingerprint Recognition
- Fingerprint Verification Competition (FVC2000)
- Fingerprint Verification Competition (FVC2002)
- Fingerprint Verification Competition (FVC2004)
- Fingerprint Verification Competition (FVC2006)
- PRIP Lab - National Tsing Hua University, Taiwan Prof. C. Chen
  
  Lindex101 -- 101 subjects x 4 images (300x300 pixels) -- missing images? try here!
  
  Rindex28 -- 28 subjects x 4 images (300x300 pixels)
Pedestrian Detection
- MIT (CBCL PEDESTRIAN DATABASE #1) -- 924 images (10/22 Mb) with 64 x 128 PPM format images for training
- INRIA Person dataset (970Mb)
  
  1208/566 positive windows (train/test) -- 64 x 128 bounding box size with 16/3 border pixels (train/test)
  
  1218/453 negatives images (train/test) -- 2592 x 1944 pixels
- Caltech Pedestrian Detection Benchmark (6Gb/5Gb)
- ETH dataset

Subjects for Seminars

Face Alignment
Face Frontalization
Gesture Recognition

Gait
3D finger printing

Typing/Keystroke biometrics/dynamics
Retina recognition
Hand geometry

Voice as biometrics

Multimodality - face+iris

Multimodality - face+fingerprint
Multimodality - iris+fingerprint
Multimodality - face+voice

Presence in Classes

name	wt.1	wt.2	#wt#	pa.1	pa.2	pa.3	pa.4	#pa#	#sem#	mean	final	mean*	situation	frequency	absences
ALAN GONZALEZ MEGER ZANGRANDI	50	30	16	70	40	40	0	14	0	30		30	rejected	65%	8/24 9/5 9/28 10/31 11/09 11/16 11/28 11/30 12/05 12/07 12/12
CAINA COSTA TREVISAN										0		0	rejected	3%	8/1 8/3 8/8 8/15 8/17 8/22 8/24 8/29 9/5 9/12 9/14 9/19 9/21 9/26 9/28 10/3 10/10 10/17 10/19 10/31 11/07 11/09 11/14 11/16 11/21 11/23 11/28 11/30 12/05 12/07 12/12
CLARISSA DREISCHERF PEREIRA	50	70	24	0	75	75	25	19	14	57	83	70	approved	84%	8/1 9/5 11/09 11/16 11/21
MATHEUS AGIO NERONE	100	70	34	100	95	80	30	29	18	81		85	approved	87%	8/29 10/17 11/09 12/07
RENAN LUCIANO BURDA	80	65	29	100	90	30	55	25	15	69		70	approved	84%	9/14 10/3 10/10 11/09 11/28
VICTOR MOCELIN	90	75	33	120	95	70	45	31	12	76		80	approved	84%	8/8 8/10 8/29 9/28 10/19
ANDERSON TADASHI SANTOS IGARASHI	75	55	26	100	95	110	75	38	19	83		85	approved	96%	11/28
CIDES SEMPREBOM BEZERRA	20	50	14	100	90	75	85	34	18	66		70	approved	96%	10/10
ERIC GAMA MÜLLER	80	90	34	140	130	110	90	46	16	96		100	approved	96%	12/12
FABRÍCIO JOSÉ DE OLIVEIRA CESCHIN	85	85	34	100	110	85	60	34	18	86		90	approved	96%	11/28
GUILHERME ZASYEKI MACHADO	85	90	35	100	100	105	75	38	18	91		95	approved	100%
JOÃO FELIPE LOPES DE SUS	50	0	10	15	40	25	0	8	0	18		20	rejected	37%	8/10 9/5 9/26 9/28 10/3 10/10 10/17 10/19 10/31 11/07 11/09 11/14 11/16 11/21 11/23 11/28 11/30 12/05 12/07 12/12
LUIZ ANTONIO ZANLORENSI JUNIOR	85	60	29	135	130	105	100	46	16	91		95	approved	93%	8/1 9/19
JEOVANE HONORIO ALVES	95	60	31	125	80	50	30	26	17	74		75	approved	81%	8/8 8/24 8/29 9/19 9/26 11/07
PEDRO MURILO DA SILVA PEREIRA	65	75	28	100	95	105	95	40	18	86		90	approved	78%	8/22 10/31 11/16 11/21 11/28 12/05 12/12

wt.1 - written test 1 (100 pts)
wt.2 - written test 2 (100 pts)
#wt# - written tests total points in 40%
pa.1 - practical activity 1 (100 pts)
pa.2 - practical activity 2 (100 pts)
pa.3 - practical activity 3 (100 pts)
pa.4 - practical activity 4 (100 pts)
#pa# - practical activites total points in 40%
#sem# - seminars total points in 20%
mean & mean* - final grad

CI396 / CI852 - Topics in Computer Vision