Deep spectral methods: A surprisingly strong baseline for unsupervised semantic segmentation and localization

L Melas-Kyriazi, C Rupprecht… - Proceedings of the …, 2022 - openaccess.thecvf.com
Unsupervised localization and segmentation are long-standing computer vision challenges
that involve decomposing an image into semantically-meaningful segments without any …

Diffuse Attend and Segment: Unsupervised Zero-Shot Segmentation using Stable Diffusion

J Tian, L Aggarwal, A Colaco, Z Kira… - Proceedings of the …, 2024 - openaccess.thecvf.com
Producing quality segmentation masks for images is a fundamental problem in computer
vision. Recent research has explored large-scale supervised training to enable zero-shot …

Railroad is not a train: Saliency as pseudo-pixel supervision for weakly supervised semantic segmentation

S Lee, M Lee, J Lee, H Shim - Proceedings of the IEEE/CVF …, 2021 - openaccess.thecvf.com
Existing studies in weakly-supervised semantic segmentation (WSSS) using image-level
weak supervision have several limitations: sparse object coverage, inaccurate object …

Unsupervised semantic segmentation by contrasting object mask proposals

W Van Gansbeke, S Vandenhende… - Proceedings of the …, 2021 - openaccess.thecvf.com
Being able to learn dense semantic representations of images without supervision is an
important problem in computer vision. However, despite its significance, this problem …

Self: Learning to filter noisy labels with self-ensembling

DT Nguyen, CK Mummadi, TPN Ngo… - arXiv preprint arXiv …, 2019 - arxiv.org
Deep neural networks (DNNs) have been shown to over-fit a dataset when being trained
with noisy labels for a long enough time. To overcome this problem, we present a simple …

Weakly-supervised salient object detection via scribble annotations

J Zhang, X Yu, A Li, P Song, B Liu… - Proceedings of the …, 2020 - openaccess.thecvf.com
Compared with laborious pixel-wise dense labeling, it is much easier to label data by
scribbles, which only costs 1 2 seconds to label one image. However, using scribble labels …

[HTML][HTML] Local contrastive loss with pseudo-label based self-training for semi-supervised medical image segmentation

K Chaitanya, E Erdil, N Karani, E Konukoglu - Medical image analysis, 2023 - Elsevier
Supervised deep learning-based methods yield accurate results for medical image
segmentation. However, they require large labeled datasets for this, and obtaining them is a …

[PDF][PDF] 基于深度学习的显著性目标检测方法综述

罗会兰, 袁璞, 童康 - 电子学报, 2021 - ejournal.org.cn
显著性目标检测旨在对图像中最显著的对象进行检测和分割, 是计算机视觉任务中重要的预处理
步骤之一, 且在信息检索, 公共安全等领域均有广泛的应用. 本文对近期基于深度学习的显著性 …

Deep unsupervised part-whole relational visual saliency

Y Liu, X Dong, D Zhang, S Xu - Neurocomputing, 2024 - Elsevier
Abstract Deep Supervised Salient Object Detection (SSOD) excessively relies on large-
scale annotated pixel-level labels which consume intensive labour acquiring high quality …

Pushing the limits of self-supervised resnets: Can we outperform supervised learning without labels on imagenet?

N Tomasev, I Bica, B McWilliams, L Buesing… - arXiv preprint arXiv …, 2022 - arxiv.org
Despite recent progress made by self-supervised methods in representation learning with
residual networks, they still underperform supervised learning on the ImageNet classification …