Evaluation of Randomization through Style Transfer for Enhanced Domain Generalization

Dustin Eisenhardt, Timothy Schaumlöffel, Alperen Kantarci, Gemma Roig · 2026

Read Evaluation of Randomization through Style Transfer for Enhanced Domain Generalization by Dustin Eisenhardt, Timothy Schaumlöffel, Alperen Kantarci, Gemma Roig (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-05616.txt.

Description

Eisenhardt and three colleagues settle conflicting advice on style transfer for domain generalization in driving scenes. Expanding the style pool beats reusing few styles, texture complexity matters little once the pool is large, and diverse artistic styles beat domain-matched ones. Their lightweight StyleMixDG recipe needs no architecture changes and improves strong baselines on GTAV-to-real benchmarks.

Catalog record

CREATORDustin Eisenhardt, Timothy Schaumlöffel, Alperen Kantarci, Gemma Roig
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25468
LICENSEhttp://creativecommons.org/licenses/by-sa/4.0/
RIGHTSCC BY SA
SOURCEarXiv

Source: https://arxiv.org/abs/2604.05616

Browse the full library · About the Tower of Babel