Is "visual rendering" the future of QoE optimization?
Is visual rendering the future abstraction for optimizing quality of experience (QoE)? A new workshop paper (2008.04128 A New Abstraction for Internet QoE Optimization) out of the University of Chicago and Microsoft Research seems to think so.
Why change the abstraction? Although user-perceived QoE is subjective by nature, Internet video streaming optimization has typically centered on optimizing a QoE model based on human engineered features. And optimizing QoE has largely plateaued, despite an ever growing list of features.
Visual rendering, on the other hand, is “a video stream that records all of the pixels displayed on the screen over time as seen by the user.” This provides two main advantages:
- expressibility: it captures a full visual experience of a user, thereby bypassing the need for human engineered features and potential abstraction mismatch with actual user perception
- generalizability: it is a unifying abstraction for all Internet applications
What sort of architecture would be required for this new approach? The paper outlines two main requirements:
- A visual renderer that infers the visual rendering of taking an optimization action.
- A QoE model that takes the visual rendering as input and predicts the QoE of a given optimization action.
It then goes on to draw upon ideas from other disciplines to suggest ideas for how QoE modeling based on visual rendering might work. From cognitive visual perception, it borrows the ideas of expectation and attention to posit high QoE as meeting exceptions without the attention region. From computer vision, visual attention detection, video summarization, and video prediction techniques could be used in implementing the new QoE model concept.
Will this be a new frontier in machine learning in networking? I guess we’ll have to wait and see.
Related
Archive
chinese
tang-dynasty-poetry
李白
python
王维
rl
pytorch
numpy
emacs
杜牧
spinningup
networking
deep-learning
贺知章
白居易
王昌龄
杜甫
李商隐
tips
reinforcement-learning
macports
jekyll
骆宾王
贾岛
孟浩然
xcode
time-series
terminal
regression
rails
productivity
pandas
math
macosx
lesson-plan
helicopters
flying
fastai
conceptual-learning
command-line
bro
黄巢
韦应物
陈子昂
王翰
王之涣
柳宗元
杜秋娘
李绅
张继
孟郊
刘禹锡
元稹
youtube
visdom
system
sungho
stylelint
stripe
softmax
siri
sgd
scipy
scikit-learn
scikit
safari
research
qtran
qoe
qmix
pyhton
poetry
pedagogy
papers
paper-review
optimization
openssl
openmpi
nyc
node
neural-net
multiprocessing
mpi
morl
ml
mdp
marl
mandarin
macos
machine-learning
latex
language-learning
khan-academy
jupyter-notebooks
ios-programming
intuition
homebrew
hacking
google-cloud
github
flashcards
faker
docker
dme
deepmind
dec-pomdp
data-wrangling
craftsman
congestion-control
coding
books
book-review
atari
anki
analogy
3brown1blue
2fa