|
Doctoral courses
-
Models for representing images and video
(fer)
-
Dynamic scene analysis
(fer)
Master courses
-
Deep Learning 1
(info,
fer)
-
Design Patterns
(info,
fer)
-
Computer architecture 2:
(info,
fer)
-
Computer vision
(info,
fer)
-
Threedimensional computer vision
(fer)
-
Scripting languages
(fer)
Mentor driven courses
Contact Address
Sinisa Segvic
University of Zagreb
Faculty of El. Eng. and Computing
(www)
Department of Electronics, Microelectronics,
Computer and Intelligent Systems
(www)
Unska 3, 10000 Zagreb, Croatia
Room: D314
Phone: +385 98 313-325
Messages: +385 1 6129-935
Fax.: +385 1 6129-653
E-mail: 
PGP: public key
Workshops
-
BMVC workshop on Robust Recognition in the Open World:
RROW 24
(in collaboration with TU Berlin and Uni Wuppertal),
Glasgow 2024.
-
EEML workshop on Machine Learning:
CMLW 2024
(in collaboration with Google DeepMind),
Zagreb 2024.
-
Workshop on Robust Understanding of Street Scenes
Using Computer Vision:
workshop home
(in collaboration with Uni Wuppertal),
Zagreb 2022.
Publications:
hr,
en,
crosbi,
scholar,
orcid,
D307@yt
Official pages:
@fer.hr (in Croatian)
Biography:
en
Curriculum vitae:
en
|
|
Research
-
Domain adaptation:
martinovic24eccv
-
Dense open-set recognition:
delic24bmvc,
workshop24,
grcic24pami,
grcic23cvprw,
kumar23cvpr,
workshop22,
grcic24sensors
grcic21visapp
bevandic22ivc
(editor's choice)
-
Defenses from data poisoning:
sabolic24bmvc
-
Automatic road-safety inspection:
kacan24arxiv,
kacan20itsc
-
Multi-domain learning with overlapping labels:
bevandic24ijcv,
bevandic22bmvc
(urcv20 best paper),
orsic20arxiv
(rvc20 winners)
-
Dense semantic forecasting
with F2MF models:
saric22tnnls,
saric20cvpr,
saric19gcpr
-
Densely connected normalizing flows:
grcic21neurips
-
One-way consistency for semi-supervised semantic segmentation:
grubisic21mva
-
Real-time Semantic Segmentation with pyramid fusion:
orsic20rvc,
orsic21pr,
orsic19cvpr
-
Ladder-style DenseNets for Semantic Segmentation
of Large Natural Images:
kreso21tits,
kreso19arxiv
kreso17cvrsuad
Alumni
-
Marin Kačan:
Recognition of road infrastructure safety attributes
by computer vision
(pdf)
-
Matej Grcić:
Synthetic negatives and unnormalized likelihood
for open-set semantic segmentation of images
(pdf,
github)
-
Ivan Grubišić:
Semi-supervised learning of semantic segmentation of images
by applying consistency criteria
(pdf,
github)
-
Petra Bevandić:
Open-set segmentation of images
by means of negative examples
(pdf,
github)
-
Josip Saric:
Joint Forecasting of Features and Feature Motion
for Semantic Future Prediction in Video
(pdf,
github)
-
Marin Orsic:
Efficient semantic image segmentation using pyramidal fusion
(pdf,
github)
-
Ivan Kreso:
Convolutional architecture for efficient semantic segmentation of large images
(pdf,
github)
-
Ivan Sikiric:
Improving fleet management systems by computer vision
(pdf,
RG)
-
Valentina Zadrija:
Bottom-up Object Localization With Fisher Vectors
(pdf,
LI)
Research projects
-
ADEPT -
Advanced dense prediction (2021-2024),
Croatian Science Foundation, project leader
-
DATACROSS -
The Centre of Research Excellence for
Data Science and Advanced Cooperative Systems.
ERDF KK.01.1.1.01.0009 (leaders: I.Petrovic, S.Loncaric)
-
MULTICLOD -
Multi-class object detection for smart cars and safe roads (2014-2017),
Croatian Science Foundation, project leader:
pr21,
tits20,
gcpr19,
iccvw17,
gcpr16a,
gcpr16b,
visapp15a,
visapp15b
|
