paper_authors: Daniel Saragih, Pascal Tyrrell for: Synthetic labeled polyp images were generated to augment automatic medical image segmentation models.methods: Diffusion models were used to generate and style synthetic labeled data.results: The generated images more closely resembled real images, as shown by lower FID scores compared to previous GAN methods. The segmentation model performed better when trained or augmented with synthetic data, as evidenced by higher DL and IoU scores.Abstract
In this paper, we proposed and evaluated a pipeline for generating synthetic labeled polyp images with the aim of augmenting automatic medical image segmentation models. In doing so, we explored the use of diffusion models to generate and style synthetic labeled data. The HyperKvasir dataset consisting of 1000 images of polyps in the human GI tract obtained from 2008 to 2016 during clinical endoscopies was used for training and testing. Furthermore, we did a qualitative expert review, and computed the Fr\'echet Inception Distance (FID) and Multi-Scale Structural Similarity (MS-SSIM) between the output images and the source images to evaluate our samples. To evaluate its augmentation potential, a segmentation model was trained with the synthetic data to compare their performance with the real data and previous Generative Adversarial Networks (GAN) methods. These models were evaluated using the Dice loss (DL) and Intersection over Union (IoU) score. Our pipeline generated images that more closely resembled real images according to the FID scores (GAN: $118.37 \pm 1.06 \text{ vs SD: } 65.99 \pm 0.37$). Improvements over GAN methods were seen on average when the segmenter was entirely trained (DL difference: $-0.0880 \pm 0.0170$, IoU difference: $0.0993 \pm 0.01493$) or augmented (DL difference: GAN $-0.1140 \pm 0.0900 \text{ vs SD }-0.1053 \pm 0.0981$, IoU difference: GAN $0.01533 \pm 0.03831 \text{ vs SD }0.0255 \pm 0.0454$) with synthetic data. Overall, we obtained more realistic synthetic images and improved segmentation model performance when fully or partially trained on synthetic data.
摘要
在这篇论文中，我们提出并评估了一个涉及扩充自动医疗图像分割模型的数据生成管道。为此，我们利用了扩散模型来生成和风格化合成数据。我们使用的是HyperKvasir dataset，包含2008-2016年期间在临床endooscopy中获取的1000个肠道肿瘤图像。我们进行了专家评审，并计算了Fréchet Inception Distance（FID）和Multi-Scale Structural Similarity（MS-SSIM）指标来评估我们的样本。为了评估它的扩充潜力，我们使用合成数据训练了一个分割模型，并与实际数据和前一代生成对抗网络（GAN）方法进行比较。这些模型被评估使用Dice损失（DL）和交集 sobre union（IoU）分数。我们的管道生成的图像更加接近实际图像，根据FID分数（GAN：$118.37 \pm 1.06 \text{ vs SD: } 65.99 \pm 0.37）。在GAN方法上 average 的情况下，我们看到了使用合成数据训练的改善（DL差异：GAN $-0.1140 \pm 0.0900 \text{ vs SD }-0.1053 \pm 0.0981$, IoU差异：GAN $0.01533 \pm 0.03831 \text{ vs SD }0.0255 \pm 0.0454$）。总的来说，我们得到了更加真实的合成图像，并使分割模型在使用合成数据进行训练或扩充时表现得更好。
results: 文献回顾了单像素成像基于深度学习的研究进展，包括超解像单像素成像、通过散射媒体的单像素成像、光子级单像素成像、基于单像素成像的光学加密、色彩单像素成像和无影像感测等领域的研究。Abstract
Since the advent of single-pixel imaging and machine learning, both fields have flourished, but followed parallel tracks. Until recently, machine learning, especially deep learning, has demonstrated effectiveness in delivering high-quality solutions across various application domains of single-pixel imaging. This article comprehensively reviews the research of single-pixel imaging technology based on deep learning. From the basic principles of single-pixel imaging and deep learning, the principles and implementation methods of single-pixel imaging based on deep learning are described in detail. Then, the research status and development trend of single-pixel imaging based on deep learning in various domains are analyzed, including super-resolution single-pixel imaging, single-pixel imaging through scattering media, photon-level single-pixel imaging, optical encryption based on single-pixel imaging, color single-pixel imaging, and image-free sensing. Finally, this review explores the limitations in the ongoing research, while offers the delivering insights into prospective avenues for future research.
摘要
(Simplified Chinese translation)自单Pixel影像和机器学习出现以来，这两个领域都在繁荣，但是在平行的轨道上进行了渐进的发展。直到最近，深度学习在各种单Pixel影像应用领域中表现出了高质量的解决方案。本文对单Pixel影像技术基于深度学习的研究进行了详细的查询。从单Pixel影像和深度学习的基本原理到深度学习在单Pixel影像方面的实现方法，都是在文中进行了详细的介绍。然后，文中分析了基于深度学习的单Pixel影像在不同领域的研究状况和发展趋势，包括超分辨单Pixel影像、单Pixel影像 через散射媒体、光子级单Pixel影像、基于单Pixel影像的光学加密、彩色单Pixel影像和无图像感知。最后，文中探讨了当前研究的限制，并提供了未来研究的可能性和挑战。
TILT: topological interface recovery in limited-angle tomography
paper_authors: Elli Karvonen, Matti Lassas, Pekka Pankka, Samuli Siltanen
for: solves the severely ill-posed inverse problem of limited-angle tomography
methods: lifting the visible part of the wavefront set under a universal covering map, using dual-tree complex wavelets, a dedicated metric, and persistent homology
results: not only a suggested invisible boundary but also a computational representation for all interfaces in the targetAbstract
A novel reconstruction method is introduced for the severely ill-posed inverse problem of limited-angle tomography. It is well known that, depending on the available measurement, angles specify a subset of the wavefront set of the unknown target, while some oriented singularities remain invisible in the data. Topological Interface recovery for Limited-angle Tomography, or TILT, is based on lifting the visible part of the wavefront set under a universal covering map. In the space provided, it is possible to connect the appropriate pieces of the lifted wavefront set correctly using dual-tree complex wavelets, a dedicated metric, and persistent homology. The result is not only a suggested invisible boundary but also a computational representation for all interfaces in the target.
摘要
新的重建方法被介绍用于受限角度 computed tomography 的强度不确定问题。已知，根据可用的测量，角度Specify a subset of the wavefront set of the unknown target, while some oriented singularities remain invisible in the data。TILT 基于提升可见部分的波front set under a universal covering map。在提供的空间中，可以正确地连接相应的部分 lifting wavefront set using dual-tree complex wavelets, a dedicated metric, and persistent homology。结果不仅是一个建议的隐藏边界，还是一个计算表示所有界面在目标中。