arXiv subjects

Hao Du

Publications and source records attributed to Hao Du.

28 records · Page 2Linked to original sources

arXiv2023

Weakly-Supervised 3D Medical Image Segmentation using Geometric Prior and Contrastive Similarity

Medical image segmentation is almost the most important pre-processing procedure in computer-aided diagnosis but is also a very challenging task due to the complex shapes of segments and various artifacts caused by medical imaging, (i.e., low-contrast tissues, and non-homogenous textures). In this paper, we propose a simple yet effective segmentation framework that incorporates the geometric prior and contrastive similarity into the weakly-supervised segmentation framework in a loss-based fashion. The proposed geometric prior built on point cloud provides meticulous geometry to the weakly-supervised segmentation proposal, which serves as better supervision than the inherent property of the bounding-box annotation (i.e., height and width). Furthermore, we propose contrastive similarity to encourage organ pixels to gather around in the contrastive embedding space, which helps better distinguish low-contrast tissues. The proposed contrastive embedding space can make up for the poor representation of the conventionally-used gray space. Extensive experiments are conducted to verify the effectiveness and the robustness of the proposed weakly-supervised segmentation framework. The proposed framework is superior to state-of-the-art weakly-supervised methods on the following publicly accessible datasets: LiTS 2017 Challenge, KiTS 2021 Challenge, and LPBA40. We also dissect our method and evaluate the performance of each component.

eess.IV↗

arXiv2022

Cyclic Differentiable Architecture Search

Differentiable ARchiTecture Search, i.e., DARTS, has drawn great attention in neural architecture search. It tries to find the optimal architecture in a shallow search network and then measures its performance in a deep evaluation network. The independent optimization of the search and evaluation networks, however, leaves room for potential improvement by allowing interaction between the two networks. To address the problematic optimization issue, we propose new joint optimization objectives and a novel Cyclic Differentiable ARchiTecture Search framework, dubbed CDARTS. Considering the structure difference, CDARTS builds a cyclic feedback mechanism between the search and evaluation networks with introspective distillation. First, the search network generates an initial architecture for evaluation, and the weights of the evaluation network are optimized. Second, the architecture weights in the search network are further optimized by the label supervision in classification, as well as the regularization from the evaluation network through feature distillation. Repeating the above cycle results in joint optimization of the search and evaluation networks and thus enables the evolution of the architecture to fit the final evaluation network. The experiments and analysis on CIFAR, ImageNet and NAS-Bench-201 demonstrate the effectiveness of the proposed approach over the state-of-the-art ones. Specifically, in the DARTS search space, we achieve 97.52% top-1 accuracy on CIFAR10 and 76.3% top-1 accuracy on ImageNet. In the chain-structured search space, we achieve 78.2% top-1 accuracy on ImageNet, which is 1.1% higher than EfficientNet-B0. Our code and models are publicly available at https://github.com/microsoft/Cream.

cs.CV↗

arXiv2021

Binomial Determinants for Tiling Problems Yield to the Holonomic Ansatz

We present and prove closed form expressions for some families of binomial determinants with signed Kronecker deltas that are located along an arbitrary diagonal in the corresponding matrix. They count cyclically symmetric rhombus tilings of hexagonal regions with triangular holes. We extend a previous systematic study of these families, where the locations of the Kronecker deltas depended on an additional parameter, to families with negative Kronecker deltas. By adapting Zeilberger's holonomic ansatz to make it work for our problems, we can take full advantage of computer algebra tools for symbolic summation. This, together with the combinatorial interpretation, allows us to realize some new determinantal relationships. From there, we are able to resolve all remaining open conjectures related to these determinants, including one from 2005 due to Lascoux and Krattenthaler.

math.CO↗

arXiv2021

Deep Portrait Lighting Enhancement with 3D Guidance

Despite recent breakthroughs in deep learning methods for image lighting enhancement, they are inferior when applied to portraits because 3D facial information is ignored in their models. To address this, we present a novel deep learning framework for portrait lighting enhancement based on 3D facial guidance. Our framework consists of two stages. In the first stage, corrected lighting parameters are predicted by a network from the input bad lighting image, with the assistance of a 3D morphable model and a differentiable renderer. Given the predicted lighting parameter, the differentiable renderer renders a face image with corrected shading and texture, which serves as the 3D guidance for learning image lighting enhancement in the second stage. To better exploit the long-range correlations between the input and the guidance, in the second stage, we design an image-to-image translation network with a novel transformer architecture, which automatically produces a lighting-enhanced result. Experimental results on the FFHQ dataset and in-the-wild images show that the proposed method outperforms state-of-the-art methods in terms of both quantitative metrics and visual quality. We will publish our dataset along with more results on https://cassiepython.github.io/egsr/index.html.

cs.CV↗

arXiv2021

Multi-task Graph Convolutional Neural Network for Calcification Morphology and Distribution Analysis in Mammograms

The morphology and distribution of microcalcifications in a cluster are the most important characteristics for radiologists to diagnose breast cancer. However, it is time-consuming and difficult for radiologists to identify these characteristics, and there also lacks of effective solutions for automatic characterization. In this study, we proposed a multi-task deep graph convolutional network (GCN) method for the automatic characterization of morphology and distribution of microcalcifications in mammograms. Our proposed method transforms morphology and distribution characterization into node and graph classification problem and learns the representations concurrently. Through extensive experiments, we demonstrate significant improvements with the proposed multi-task GCN comparing to the baselines. Moreover, the achieved improvements can be related to and enhance clinical understandings. We explore, for the first time, the application of GCNs in microcalcification characterization that suggests the potential of graph learning for more robust understanding of medical images.

cs.CV↗

arXiv2021

Cream of the Crop: Distilling Prioritized Paths For One-Shot Neural Architecture Search

One-shot weight sharing methods have recently drawn great attention in neural architecture search due to high efficiency and competitive performance. However, weight sharing across models has an inherent deficiency, i.e., insufficient training of subnetworks in hypernetworks. To alleviate this problem, we present a simple yet effective architecture distillation method. The central idea is that subnetworks can learn collaboratively and teach each other throughout the training process, aiming to boost the convergence of individual models. We introduce the concept of prioritized path, which refers to the architecture candidates exhibiting superior performance during training. Distilling knowledge from the prioritized paths is able to boost the training of subnetworks. Since the prioritized paths are changed on the fly depending on their performance and complexity, the final obtained paths are the cream of the crop. We directly select the most promising one from the prioritized paths as the final architecture, without using other complex search methods, such as reinforcement learning or evolution algorithms. The experiments on ImageNet verify such path distillation method can improve the convergence ratio and performance of the hypernetwork, as well as boosting the training of subnetworks. The discovered architectures achieve superior performance compared to the recent MobileNetV3 and EfficientNet families under aligned settings. Moreover, the experiments on object detection and more challenging search space show the generality and robustness of the proposed method. Code and models are available at https://github.com/microsoft/cream.git.

cs.CV↗

arXiv2020

An Additive Decomposition in S-Primitive Towers

We consider the additive decomposition problem in primitive towers and present an algorithm to decompose a function in an S-primitive tower as a sum of a derivative in the tower and a remainder which is minimal in some sense. Special instances of S-primitive towers include differential fields generated by finitely many logarithmic functions and logarithmic integrals. A function in an S-primitive tower is integrable in the tower if and only if the remainder is equal to zero. The additive decomposition is achieved by viewing our towers not as a traditional chain of extension fields, but rather as a direct sum of certain subrings. Furthermore, we can determine whether or not a function in an S-primitive tower has an elementary integral without solving any differential equations. We also show that a kind of S-primitive towers, known as logarithmic towers, can be embedded into a particular extension where we can obtain a finer remainder.

cs.SC↗

arXiv2019

Zoom in to where it matters: a hierarchical graph based model for mammogram analysis

In clinical practice, human radiologists actually review medical images with high resolution monitors and zoom into region of interests (ROIs) for a close-up examination. Inspired by this observation, we propose a hierarchical graph neural network to detect abnormal lesions from medical images by automatically zooming into ROIs. We focus on mammogram analysis for breast cancer diagnosis for this study. Our proposed network consist of two graph attention networks performing two tasks: (1) node classification to predict whether to zoom into next level; (2) graph classification to classify whether a mammogram is normal/benign or malignant. The model is trained and evaluated on INbreast dataset and we obtain comparable AUC with state-of-the-art methods.

eess.IV↗

arXiv2019

A Self-Correcting Deep Learning Approach to Predict Acute Conditions in Critical Care

In critical care, intensivists are required to continuously monitor high dimensional vital signs and lab measurements to detect and diagnose acute patient conditions. This has always been a challenging task. In this study, we propose a novel self-correcting deep learning prediction approach to address this challenge. We focus on an example of the prediction of acute kidney injury (AKI). Compared with the existing models, our method has a number of distinct features: we utilized the accumulative data of patients in ICU; we developed a self-correcting mechanism that feeds errors from the previous predictions back into the network; we also proposed a regularization method that takes into account not only the model's prediction error on the label but also its estimation errors on the input data. This mechanism is applied in both regression and classification tasks. We compared the performance of our proposed method with the conventional deep learning models on two real-world clinical datasets and demonstrated that our proposed model constantly outperforms these baseline models. In particular, the proposed model achieved area under ROC curve at 0.893 on the MIMIC III dataset, and 0.871 on the Philips eICU dataset.

cs.LG↗

arXiv2018

Additive Decompositions in Primitive Extensions

This paper extends the classical Ostrogradsky-Hermite reduction for rational functions to more general functions in primitive extensions of certain types. For an element $f$ in such an extension $K$, the extended reduction decomposes $f$ as the sum of a derivative in $K$ and another element $r$ such that $f$ has an antiderivative in $K$ if and only if $r=0$; and $f$ has an elementary antiderivative over $K$ if and only if $r$ is a linear combination of logarithmic derivatives over the constants when $K$ is a logarithmic extension. Moreover, $r$ is minimal in some sense. Additive decompositions may lead to reduction-based creative-telescoping methods for nested logarithmic functions, which are not necessarily $D$-finite.

cs.SC↗