OASIS
Online Adaptation for Semantic Image Segmentation
Original Scene
x 0.1
x 0.25
x 0.5
x 1
x 2
Categories
all
bicycle
bus
car
moto
person
rider
train
truck
building
fence
foliage
road
sidewalk
sky
terrain
wall
light
pole
sign
No Adaptation
Naive Continual Learning
NLE Continual Learning
© 2022 NAVER LABS Europe