Searcharxiv⌕ Search

arXiv subjects

Yongbing Zhang

Publications and source records attributed to Yongbing Zhang.

At least 73 records · Page 4Linked to original sources
arXiv2022

WSSS4LUAD: Grand Challenge on Weakly-supervised Tissue Semantic Segmentation for Lung Adenocarcinoma

Lung cancer is the leading cause of cancer death worldwide, and adenocarcinoma (LUAD) is the most common subtype. Exploiting the potential value of the histopathology images can promote precision medicine in oncology. Tissue segmentation is the basic upstream task of histopathology image analysis. Existing deep learning models have achieved superior segmentation performance but require sufficient pixel-level annotations, which is time-consuming and expensive. To enrich the label resources of LUAD and to alleviate the annotation efforts, we organize this challenge WSSS4LUAD to call for the outstanding weakly-supervised semantic segmentation (WSSS) techniques for histopathology images of LUAD. Participants have to design the algorithm to segment tumor epithelial, tumor-associated stroma and normal tissue with only patch-level labels. This challenge includes 10,091 patch-level annotations (the training set) and over 130 million labeled pixels (the validation and test sets), from 87 WSIs (67 from GDPH, 20 from TCGA). All the labels were generated by a pathologist-in-the-loop pipeline with the help of AI models and checked by the label review board. Among 532 registrations, 28 teams submitted the results in the test phase with over 1,000 submissions. Finally, the first place team achieved mIoU of 0.8413 (tumor: 0.8389, stroma: 0.7931, normal: 0.8919). According to the technical reports of the top-tier teams, CAM is still the most popular approach in WSSS. Cutmix data augmentation has been widely adopted to generate more reliable samples. With the success of this challenge, we believe that WSSS approaches with patch-level annotations can be a complement to the traditional pixel annotations while reducing the annotation efforts. The entire dataset has been released to encourage more researches on computational pathology in LUAD and more novel WSSS techniques.

eess.IV↗
arXiv2022

Lensless coherent diffraction imaging based on spatial light modulator with unknown modulation curve

Lensless imaging is a popular research field for the advantages of small size, wide field-of-view and low aberration in recent years. However, some traditional lensless imaging methods suffer from slow convergence, mechanical errors and conjugate solution interference, which limit its further application and development. In this work, we proposed a lensless imaging method based on spatial light modulator (SLM) with unknown modulation curve. In our imaging system, we use SLM to modulate the wavefront of object, and introduce the ptychographic scanning algorithm that is able to recover the complex amplitude information even the SLM modulation curve is inaccurate or unknown. In addition, we also design a split-beam interference experiment to calibrate the modulation curve of SLM, and using the calibrated modulation function as the initial value of the expended ptychography iterative engine (ePIE) algorithm can improve the convergence speed. We further analyze the effect of modulation function, algorithm parameters and the characteristics of the coherent light source on the quality of reconstructed image. The simulated and real experiments show that the proposed method is superior to traditional mechanical scanning methods in terms of recovering speed and accuracy, with the recovering resolution up to 14 um.

physics.optics↗
arXiv2021

TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classification

Multiple instance learning (MIL) is a powerful tool to solve the weakly supervised classification in whole slide image (WSI) based pathology diagnosis. However, the current MIL methods are usually based on independent and identical distribution hypothesis, thus neglect the correlation among different instances. To address this problem, we proposed a new framework, called correlated MIL, and provided a proof for convergence. Based on this framework, we devised a Transformer based MIL (TransMIL), which explored both morphological and spatial information. The proposed TransMIL can effectively deal with unbalanced/balanced and binary/multiple classification with great visualization and interpretability. We conducted various experiments for three different computational pathology problems and achieved better performance and faster convergence compared with state-of-the-art methods. The test AUC for the binary tumor classification can be up to 93.09% over CAMELYON16 dataset. And the AUC over the cancer subtypes classification can be up to 96.03% and 98.82% over TCGA-NSCLC dataset and TCGA-RCC dataset, respectively. Implementation is available at: https://github.com/szc19990412/TransMIL.

cs.CV↗
arXiv2021

Deep-3D Microscope: 3D volumetric microscopy of thick scattering samples using a wide-field microscope and machine learning

Confocal microscopy is the standard approach for obtaining volumetric images of a sample with high axial and lateral resolution, especially when dealing with scattering samples. Unfortunately, a confocal microscope is quite expensive compared to traditional microscopes. In addition, the point scanning in a confocal leads to slow imaging speed and photobleaching due to the high dose of laser energy. In this paper, we demonstrate how the advances in machine learning can be exploited to "teach" a traditional wide-field microscope, one that's available in every lab, into producing 3D volumetric images like a confocal. The key idea is to obtain multiple images with different focus settings using a wide-field microscope and use a 3D Generative Adversarial Network (GAN) based neural network to learn the mapping between the blurry low-contrast image stack obtained using wide-field and the sharp, high-contrast images obtained using a confocal. After training the network with widefield-confocal image pairs, the network can reliably and accurately reconstruct 3D volumetric images that rival confocal in terms of its lateral resolution, z-sectioning and image contrast. Our experimental results demonstrate generalization ability to handle unseen data, stability in the reconstruction results, high spatial resolution even when imaging thick ($\sim40$ microns) highly-scattering samples. We believe that such learning-based-microscopes have the potential to bring confocal quality imaging to every lab that has a wide-field microscope.

physics.optics↗
arXiv2020

Freshness-Optimal Caching for Information Updating Systems with Limited Cache Storage Capacity

In this paper, we investigate a cache updating system with a server containing $N$ files, $K$ relays and $M$ users. The server keeps the freshest versions of the files which are updated with fixed rates. Each relay can download the fresh files from the server in a certain period of time. Each user can get the fresh files from any relay as long as the relay has stored the fresh versions of the requested files. Due to the limited storage capacity and updating capacity of each relay, different cache designs will lead to different average freshness of all updating files at users. In order to keep the average freshness as large as possible in the cache updating system, we formulate an average freshness-optimal cache updating problem (AFOCUP) to obtain an optimal cache scheme. However, because of the nonlinearity of the AFOCUP, it is difficult to seek out the optimal cache scheme. As a result, an linear approximate model is suggested by distributing the total update rates completely in accordance with the number of files in the relay in advance. Then we utilize the greedy algorithm to search the optimal cache scheme that is satisfied with the limited storage capacity of each relay. Finally, some numerical examples are provided to illustrate the performance of the approximate solution.

cs.IT↗
arXiv2020

Depth image denoising using nuclear norm and learning graph model

The depth images denoising are increasingly becoming the hot research topic nowadays because they reflect the three-dimensional (3D) scene and can be applied in various fields of computer vision. But the depth images obtained from depth camera usually contain stains such as noise, which greatly impairs the performance of depth related applications. In this paper, considering that group-based image restoration methods are more effective in gathering the similarity among patches, a group based nuclear norm and learning graph (GNNLG) model was proposed. For each patch, we find and group the most similar patches within a searching window. The intrinsic low-rank property of the grouped patches is exploited in our model. In addition, we studied the manifold learning method and devised an effective optimized learning strategy to obtain the graph Laplacian matrix, which reflects the topological structure of image, to further impose the smoothing priors to the denoised depth image. To achieve fast speed and high convergence, the alternating direction method of multipliers (ADMM) is proposed to solve our GNNLG. The experimental results show that the proposed method is superior to other current state-of-the-art denoising methods in both subjective and objective criterion.

cs.CV↗
arXiv2019

OpenWSI: a low-cost, high-throughput whole slide imaging system via single-frame autofocusing and open-source hardware

Recent advancements in whole slide imaging (WSI) have moved pathology closer to digital practice. Existing systems require precise mechanical control and the cost is prohibitive for most individual pathologists. Here we report a low-cost and high-throughput WSI system termed OpenWSI. The reported system is built using off-the-shelf components including a programmable LED array, a photographic lens, and a low-cost computer numerical control (CNC) router. Different from conventional WSI platforms, our system performs real-time single-frame autofocusing using color-multiplexed illumination. For axial positioning control, we perform coarse adjustment using the CNC router and precise adjustment using the ultrasonic motor ring in the photographic lens. By using a 20X objective lens, we show that the OpenWSI system has a resolution of ~0.7 microns. It can acquire whole slide images of a 225-mm^2 region in ~2 mins, with throughput comparable to existing high-end platforms. The reported system offers a turnkey solution to transform the high-end WSI platforms into one that can be made broadly available and utilizable without loss of capacity.

eess.IV↗
arXiv2019

PgNN: Physics-guided Neural Network for Fourier Ptychographic Microscopy

Fourier ptychography (FP) is a newly developed computational imaging approach that achieves both high resolution and wide field of view by stitching a series of low-resolution images captured under angle-varied illumination. So far, many supervised data-driven models have been applied to solve inverse imaging problems. These models need massive amounts of data to train, and are limited by the dataset characteristics. In FP problems, generic datasets are always scarce, and the optical aberration varies greatly under different acquisition conditions. To address these dilemmas, we model the forward physical imaging process as an interpretable physics-guided neural network (PgNN), where the reconstructed image in the complex domain is considered as the learnable parameters of the neural network. Since the optimal parameters of the PgNN can be derived by minimizing the difference between the model-generated images and real captured angle-varied images corresponding to the same scene, the proposed PgNN can get rid of the problem of massive training data as in traditional supervised methods. Applying the alternate updating mechanism and the total variation regularization, PgNN can flexibly reconstruct images with improved performance. In addition, the Zernike mode is incorporated to compensate for optical aberrations to enhance the robustness of FP reconstructions. As a demonstration, we show our method can reconstruct images with smooth performance and detailed information in both simulated and experimental datasets. In particular, when validated in an extension of a high-defocus, high-exposure tissue section dataset, PgNN outperforms traditional FP methods with fewer artifacts and distinguishable structures.

eess.IV↗
arXiv2019

Rapid and robust whole slide imaging based on LED-array illumination and color-multiplexed single-shot autofocusing

Background: The use of whole slide imaging (WSI) for digital pathology has recently been cleared for primary diagnosis in the US. A conventional WSI system scans the tissue slide to different positions and acquires the digital images. In a typical implementation, a focus map is created prior to the scanning process, leading to significant overhead time and a necessity for high positional accuracy of the mechanical system. The resulting cost of WSI system is often prohibitive for frozen section procedure during surgery. Methods: We report a novel WSI scheme based on a programmable LED array for sample illumination. In between two regular brightfield image acquisitions, we acquire one additional image by turning on a red and a green LED for color multiplexed illumination. We then identify the translational shift of the red- and green-channel images by maximizing the image mutual information or cross-correlation. The resulting translational shift is used for dynamic focus correction in the scanning process. Since we track the differential focus during adjacent acquisitions, there is no positional repeatability requirement in our scheme. Results: We demonstrate a prototype WSI platform with a mean focusing error of ~0.3 microns. Different from previous implementations, this prototype platform requires no focus map surveying, no secondary camera or additional optics, and allows for continuous sample motion in the focus tracking process. Conclusions: A programmable LED array can be used for color-multiplexed single-shot autofocusing in WSI. The reported scheme may enable the development of cost-effective WSI platforms without positional repeatability requirement. It may also provide a turnkey solution for other high-content microscopy applications.

eess.IV↗
arXiv2019

Full-field Fourier ptychography (FFP): spatially varying pupil modeling and its application for rapid field-dependent aberration metrology

Digital aberration measurement and removal play a prominent role in computational imaging platforms aimed at achieving simple and compact optical arrangements. A recent important class of such platforms is Fourier ptychography, which is geared towards efficiently creating gigapixel images with high resolution and large field of view (FOV). In current FP implementations, pupil aberration is often recovered at each small segment of the entire FOV. This reconstruction strategy fails to consider the field-dependent nature of the optical pupil. Given the power series expansion of the wavefront aberration, the spatially varying pupil can be fully characterized by tens of coefficients over the entire FOV. With this observation, we report a Full-field Fourier Ptychography (FFP) scheme for rapid and robust aberration metrology. The meaning of 'full-field' in FFP is referred to the recovering of the 'full-field' coefficients that govern the field-dependent pupil over the entire FOV. The optimization degrees of freedom are at least two orders of magnitude lower than the previous implementations. We show that the image acquisition process of FFP can be completed in ~1s and the spatially varying aberration of the entire FOV can be recovered in ~35s using a CPU. The reported approach may facilitate the further development of Fourier ptychography. Since no moving part or calibration target is needed in this approach, it may find important applications in aberration metrology. The derivation of the full-field coefficients and its extension for Zernike modes also provide a general tool for analyzing spatially varying aberrations in computational imaging systems.

physics.optics↗
arXiv2019

On the Mathematical Understanding of ResNet with Feynman Path Integral

In this paper, we aim to understand Residual Network (ResNet) in a scientifically sound way by providing a bridge between ResNet and Feynman path integral. In particular, we prove that the effect of residual block is equivalent to partial differential equation, and the ResNet transforming process can be equivalently converted to Feynman path integral. These conclusions greatly help us mathematically understand the advantage of ResNet in addressing the gradient vanishing issue. More importantly, our analyses offer a path integral view of ResNet, and demonstrate that the output of certain network can be obtained by adding contributions of all paths. Moreover, the contribution of each path is proportional to e^{-S}, where S is the action given by time integral of Lagrangian L. This lays the solid foundation in the understanding of ResNet, and provides insights in the future design of convolutional neural network architecture. Based on these results, we have designed the network using partial differential operators, which further validates our theoritical analyses.

cs.LG↗
arXiv2019

Near-field Fourier ptychography: super-resolution phase retrieval via speckle illumination

Achieving high spatial resolution is the goal of many imaging systems. Designing a high-resolution lens with diffraction-limited performance over a large field of view remains a difficult task in imaging system design. On the other hand, creating a complex speckle pattern with wavelength-limited spatial features is effortless and can be implemented via a simple random diffuser. With this observation and inspired by the concept of near-field ptychography, we report a new imaging modality, termed near-field Fourier ptychography, for tackling high-resolution imaging challenges in both microscopic and macroscopic imaging settings. The meaning of 'near-field' is referred to placing the object at a short defocus distance with a large Fresnel number. In our implementations, we project a speckle pattern with fine spatial features on the object instead of directly resolving the spatial features via a high-resolution lens. We then translate the object (or speckle) to different positions and acquire the corresponding images using a low-resolution lens. A ptychographic phase retrieval process is used to recover the complex object, the unknown speckle pattern, and the coherent transfer function at the same time. In a microscopic imaging setup, we use a 0.12 numerical aperture (NA) lens to achieve a NA of 0.85 in the reconstruction process. In a macroscale photographic imaging setup, we achieve ~7-fold resolution gain using a photographic lens. The final achievable resolution is not determined by the collection optics. Instead, it is determined by the feature size of the speckle pattern. The reported imaging modality can be employed in light, coherent X-ray, and transmission electron imaging systems to increase resolution and provide quantitative absorption and phase contrast of the object.

physics.optics↗
arXiv2018

Unsupervised Image Super-Resolution using Cycle-in-Cycle Generative Adversarial Networks

We consider the single image super-resolution problem in a more general case that the low-/high-resolution pairs and the down-sampling process are unavailable. Different from traditional super-resolution formulation, the low-resolution input is further degraded by noises and blurring. This complicated setting makes supervised learning and accurate kernel estimation impossible. To solve this problem, we resort to unsupervised learning without paired data, inspired by the recent successful image-to-image translation applications. With generative adversarial networks (GAN) as the basic component, we propose a Cycle-in-Cycle network structure to tackle the problem within three steps. First, the noisy and blurry input is mapped to a noise-free low-resolution space. Then the intermediate image is up-sampled with a pre-trained deep model. Finally, we fine-tune the two modules in an end-to-end manner to get the high-resolution output. Experiments on NTIRE2018 datasets demonstrate that the proposed unsupervised method achieves comparable results as the state-of-the-art supervised models.

cs.CV↗
arXiv2017

Invariant surface area functionals and singular Yamabe problem in 3-dimensional CR geometry

We express two CR invariant surface area elements in terms of quantities in pseudohermitian geometry. We deduce the Euler-Lagrange equations of the associated energy functionals. Many solutions are given and discussed. In relation to the singular CR Yamabe problem, we show that one of the energy functionals appears as the coefficient (up to a constant multiple) of the log term in the associated volume renormalization.

math.DG↗
arXiv2014

Stabilities of homothetically shrinking Yang-Mills solitons

In this paper we introduce entropy-stability and F-stability for homothetically shrinking Yang-Mills solitons, employing entropy and second variation of $\mathcal{F}$-functional respectively. For a homothetically shrinking soliton which does not descend, we prove that entropy-stability implies F-stability. These stabilities have connections with the study of Type-I singularities of the Yang-Mills flow. Two byproducts are also included: We show that the Yang-Mills flow in dimension four cannot develop a Type-I singularity; and we obtain a gap theorem for homothetically shrinking solitons.

math.DG↗
arXiv2013

Lagrangian F-stability of closed Lagrangian self-shrinkers

In this paper, we study the Lagrangian F-stability of closed Lagrangian self-shrinkers immersed in complex Euclidean space. We show that any closed Lagrangian self-shrinker with first Betti number greater than one is Lagrangian F-unstable. In particular, any two-dimensional embedded closed Lagrangian self-shrinker is Lagrangian F-unstable. For a closed Lagrangian self-shrinker with first Betti number equal to one, we show that Lagrangian F-stability is equivalent to Hamiltonian F-stability. We also characterize Hamiltonian F-stability of a closed Lagrangian self-shrinker by its spectral property of the twisted Laplacian.

math.DG↗
arXiv2013

The Sasaki-Ricci flow on Sasakian 3-spheres

We show that on a Sasakian 3-sphere the Sasaki-Ricci flow initiating from a Sasakian metric of positive transverse scalar curvature converges to a gradient Sasaki- Ricci soliton. We also show the existence and uniqueness of gradient Sasaki-Ricci soliton on each Sasakian 3-sphere.

math.DG↗