----------------------------------------------------------------------
KU-POD dataset for 

Yeong Jun Koh and Chang-Su Kim, "Unsupervised Primary Object Discovery
in Videos Based on Evolutionary Primary Object Modeling with Reliable 
Object Proposals" submitted to IEEE TIP

If you have any comment, suggestion, or question, please contact
Yeong Jun Koh (yjkoh@mcl.korea.ac.kr).


----------------------------------------------------------------------
KU-POD dataset
----------------------------------------------------------------------
KU-POD is composed of 83 videos: 9 from Segtrack v2, 37 from FBMS, 
and 37 from VSB100.

A. Segtrack v2 [1]
http://web.engr.oregonstate.edu/~lif/SegTrack2/dataset.html

B. FBMS [2]
http://lmb.informatik.uni-freiburg.de/resources/datasets/moseg.en.html

C. VSB100 [3]
http://lmb.informatik.uni-freiburg.de/resources/datasets/vsb.en.html

See our project page for discovery results and ground-truth:
https://mcl.korea.ac.kr/PrimaryObjectDiscovery


----------------------------------------------------------------------
Sequence List
----------------------------------------------------------------------
A. Segtrack v2
bird_of_paradise
birdfall
frog
girl
monkey
monkeydog
parachute
soldier
worm

B. FBMS
  1) Test set
camel01
cars1
cars4
cats01
cats03
cats06
dogs01
dogs02
horses02
lion01
marple12
marple2
marple4
marple6
people03
people1
people2
rabbits04
tennis

  2) Training set
bear01
bear02
cars6
cars7
cars8
cats02
cats05
horses01
horses03
marple1
marple11
marple13
marple5
marple8
meerkats01
people05
rabbits01
rabbits05

C. VSB100
  1) Test set
animal_chase
buck
chameleons
fish_underwater
freight_train
gokart
harley_davidson
horse_gate
jungle_cat
kim_yu_na
koala
nordic_skiing
octopus
palm_tree
panda
planet_earth_1
rock_climbing
shark_attack
snow_leopards
snowboarding
swimming
yosemite

  2) Training set
alec_baldwin
bowling
car_jump
drone
gray_squirrel
horse_riding
lion
lion2
rock_climbing
sailing
sea_snake
sitting_dog
swing
tarantula
trampoline


----------------------------------------------------------------------
Reference
----------------------------------------------------------------------
[1] F. Li, T. Kim, A. Humayun, D. Tsai, and J. Rehg, “Video segmentation
    by tracking many figure-ground segments,” in ICCV, 2013, pp. 2192–
    2199.
[2] T. Brox and J. Malik, “Object segmentation by long term analysis of
    point trajectories,” in ECCV, 2010, pp. 282–295.
[3] F. Galasso, N. Nagaraja, T. Cardenas, T. Brox, and B. Schiele, 
    “A unified 	video segmentation benchmark: Annotation, metrics and 
    analysis,” in ICCV, 2013, pp. 3527–3534.