Exploring Simple Open-Vocabulary Semantic Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Lai, Zihang
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909079578869760
author Lai, Zihang
author_facet Lai, Zihang
contents Open-vocabulary semantic segmentation models aim to accurately assign a semantic label to each pixel in an image from a set of arbitrary open-vocabulary texts. In order to learn such pixel-level alignment, current approaches typically rely on a combination of (i) image-level VL model (e.g. CLIP), (ii) ground truth masks, and (iii) custom grouping encoders. In this paper, we introduce S-Seg, a novel model that can achieve surprisingly strong performance without depending on any of the above elements. S-Seg leverages pseudo-mask and language to train a MaskFormer, and can be easily trained from publicly available image-text datasets. Contrary to prior works, our model directly trains for pixel-level features and language alignment. Once trained, S-Seg generalizes well to multiple testing datasets without requiring fine-tuning. In addition, S-Seg has the extra benefits of scalability with data and consistently improvement when augmented with self-training. We believe that our simple yet effective approach will serve as a solid baseline for future research.
format Preprint
id arxiv_https___arxiv_org_abs_2401_12217
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Exploring Simple Open-Vocabulary Semantic Segmentation
Lai, Zihang
Computer Vision and Pattern Recognition
Machine Learning
Open-vocabulary semantic segmentation models aim to accurately assign a semantic label to each pixel in an image from a set of arbitrary open-vocabulary texts. In order to learn such pixel-level alignment, current approaches typically rely on a combination of (i) image-level VL model (e.g. CLIP), (ii) ground truth masks, and (iii) custom grouping encoders. In this paper, we introduce S-Seg, a novel model that can achieve surprisingly strong performance without depending on any of the above elements. S-Seg leverages pseudo-mask and language to train a MaskFormer, and can be easily trained from publicly available image-text datasets. Contrary to prior works, our model directly trains for pixel-level features and language alignment. Once trained, S-Seg generalizes well to multiple testing datasets without requiring fine-tuning. In addition, S-Seg has the extra benefits of scalability with data and consistently improvement when augmented with self-training. We believe that our simple yet effective approach will serve as a solid baseline for future research.
title Exploring Simple Open-Vocabulary Semantic Segmentation
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2401.12217