A Simple Recipe for Language-guided Domain Generalized Segmentation

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Fahes, Mohammad, Vu, Tuan-Hung, Bursuc, Andrei, Pérez, Patrick, de Charette, Raoul
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914738262245376
author Fahes, Mohammad
Vu, Tuan-Hung
Bursuc, Andrei
Pérez, Patrick
de Charette, Raoul
author_facet Fahes, Mohammad
Vu, Tuan-Hung
Bursuc, Andrei
Pérez, Patrick
de Charette, Raoul
contents Generalization to new domains not seen during training is one of the long-standing challenges in deploying neural networks in real-world applications. Existing generalization techniques either necessitate external images for augmentation, and/or aim at learning invariant representations by imposing various alignment constraints. Large-scale pretraining has recently shown promising generalization capabilities, along with the potential of binding different modalities. For instance, the advent of vision-language models like CLIP has opened the doorway for vision models to exploit the textual modality. In this paper, we introduce a simple framework for generalizing semantic segmentation networks by employing language as the source of randomization. Our recipe comprises three key ingredients: (i) the preservation of the intrinsic CLIP robustness through minimal fine-tuning, (ii) language-driven local style augmentation, and (iii) randomization by locally mixing the source and augmented styles during training. Extensive experiments report state-of-the-art results on various generalization benchmarks. Code is accessible at https://github.com/astra-vision/FAMix .
format Preprint
id arxiv_https___arxiv_org_abs_2311_17922
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle A Simple Recipe for Language-guided Domain Generalized Segmentation
Fahes, Mohammad
Vu, Tuan-Hung
Bursuc, Andrei
Pérez, Patrick
de Charette, Raoul
Computer Vision and Pattern Recognition
Generalization to new domains not seen during training is one of the long-standing challenges in deploying neural networks in real-world applications. Existing generalization techniques either necessitate external images for augmentation, and/or aim at learning invariant representations by imposing various alignment constraints. Large-scale pretraining has recently shown promising generalization capabilities, along with the potential of binding different modalities. For instance, the advent of vision-language models like CLIP has opened the doorway for vision models to exploit the textual modality. In this paper, we introduce a simple framework for generalizing semantic segmentation networks by employing language as the source of randomization. Our recipe comprises three key ingredients: (i) the preservation of the intrinsic CLIP robustness through minimal fine-tuning, (ii) language-driven local style augmentation, and (iii) randomization by locally mixing the source and augmented styles during training. Extensive experiments report state-of-the-art results on various generalization benchmarks. Code is accessible at https://github.com/astra-vision/FAMix .
title A Simple Recipe for Language-guided Domain Generalized Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2311.17922