# TextureSAM: Towards a Texture Aware Foundation Model for Segmentation

Source: [TextureSAM: Towards a Texture Aware Foundation Model for Segmentation](<https://www.sota2.com/research/paper/68fa78bf1e209131cea7a3a9>)

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

Data updated: 2025-10-28T13:57:16.833Z

Authors: Inbal Cohen, Boaz Meivar, Peihan Tu, Shai Avidan, Gal Oren

Published: 2025-05-22T11:31:56.000Z

## Abstract

Segment Anything Models (SAM) have achieved remarkable success in object segmentation tasks across diverse datasets. However, these models are predominantly trained on large-scale semantic segmentation datasets, which introduce a bias toward object shape rather than texture cues in the image. This limitation is critical in domains such as medical imaging, material classification, and remote sensing, where texture changes define object boundaries. In this study, we investigate SAM's bias toward semantics over textures and introduce a new texture-aware foundation model, TextureSAM, which performs superior segmentation in texture-dominant scenarios. To achieve this, we employ a novel fine-tuning approach that incorporates texture augmentation techniques, incrementally modifying training images to emphasize texture features. By leveraging a novel texture-alternation of the ADE20K dataset, we guide TextureSAM to prioritize texture-defined regions, thereby mitigating the inherent shape bias present in the original SAM model. Our extensive experiments demonstrate that TextureSAM significantly outperforms SAM-2 on both natural (+0.2 mIoU) and synthetic (+0.18 mIoU) texture-based segmentation datasets. The code and texture-augmented dataset will be publicly available.

## 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.

- [Texture Segmentation on RWTD cross-domain](<https://www.sota2.com/research/sota/texture-segmentation-on-rwtd-cross-domain.md>): Texture Segmentation; RWTD cross-domain; mIoU: 46.84.

- [Texture Segmentation on STLD (matched-domain)](<https://www.sota2.com/research/sota/texture-segmentation-on-stld-matched-domain.md>): Texture Segmentation; STLD (matched-domain); mIoU: 46.77.

- [Texture Segmentation on TextureADE](<https://www.sota2.com/research/sota/texture-segmentation-on-textureade.md>): Texture Segmentation; TextureADE; mIoU: 47.98.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2505.16540>)

[PDF](<https://arxiv.org/pdf/2505.16540.pdf>)