# AttEntropy: On the Generalization Ability of Supervised Semantic Segmentation Transformers to New Objects in New Domains

Source: [AttEntropy: On the Generalization Ability of Supervised Semantic Segmentation Transformers to New Objects in New Domains](<https://www.sota2.com/research/paper/68fa76721e209131cea6d0a8>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T03:54:27.044Z

Authors: Krzysztof Lis, Matthias Rottmann, Annika M\\"utze, Sina Honari, Pascal Fua, Mathieu Salzmann

Published: 2022-12-29T18:07:56.000Z

## Abstract

In addition to impressive performance, vision transformers have demonstrated remarkable abilities to encode information they were not trained to extract. For example, this information can be used to perform segmentation or single-view depth estimation even though the networks were only trained for image recognition. We show that a similar phenomenon occurs when explicitly training transformers for semantic segmentation in a supervised manner for a set of categories: Once trained, they provide valuable information even about categories absent from the training set. This information can be used to segment objects from these never-seen-before classes in domains as varied as road obstacles, aircraft parked at a terminal, lunar rocks, and maritime hazards.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2212.14397>)

[PDF](<https://arxiv.org/pdf/2212.14397.pdf>)