AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

Domain Adaptation Through Task Distillation

Brady Zhou, Nimit Kalra, Philipp Krähenbühl

ECCV 2020 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Geometric alignment and synthetic-to-real adaptation · 15 citations (OpenAlex)

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use map-view instead of camera-view proxytask distillation · SuperTuxKart to CARLAavg distance traveled (m): 164.6 -> 260.5Improved
use map-view instead of camera-view proxymodular pipeline · SuperTuxKart to CARLAavg distance traveled (m): 110.4 -> 135.3No clear effect
use task distillation instead of modular transfer with map-viewmodular pipeline · SuperTuxKart to CARLAavg distance traveled (m): 135.3 -> 260.5Improved
use task distillation instead of modular transfer with map-viewmodular pipeline · ViZDoom to CARLAavg distance traveled (m): 145.3 -> 277.3Improved
use task distillation instead of modular transfer with camera-viewmodular pipeline · ViZDoom to CARLAavg distance traveled (m): 140.4 -> 166.9Improved
train driving policy on ground-truth proxy labels instead of predicted proxy labels (ablation)modular pipeline, map-view · SuperTuxKart to CARLAavg distance traveled (m): 49.9 -> 135.3Improved
train driving policy on ground-truth proxy labels instead of predicted proxy labels (ablation)modular pipeline, cam-view · ViZDoom to CARLAavg distance traveled (m): 148.6 -> 140.4Hurt
transfer waypoints instead of raw steering and throttle (ablation)CyCADA · SuperTuxKart to CARLAavg distance traveled (m): 26.7 -> 24.0No clear effect
add task distillation via depth proxy to direct transferDeepLabv3 · SYNTHIA-SF to CARLAmIoU: 21.2 -> 55.2Improved
add task distillation via depth proxy to direct transferDeepLabv3 · SYNTHIA-SF to CityscapesmIoU: 15.0 -> 28.0Improved
use task distillation with DeepLabv3 instead of AT/DTAT/DT · SYNTHIA-SF to CARLAmIoU: 43.1 -> 55.2Improved
use task distillation instead of AT/DTAT/DT · SYNTHIA-SF to CityscapesmIoU: 23.2 -> 28.0Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy