PyTorchKR Papers

최신 논문

총 4,951편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2003.10690

Organ Segmentation From Full-size CT Images Using Memory-Efficient FCN

In this work, we present a memory-efficient fully convolutional network (FCN) incorporated with several memory-optimized techniques to reduce the run-time GPU memory demand during training phase. In medical image segmentation tasks, subvolume cropping has become a common preprocessing. Subvolumes (or small patch volumes) were cropped to reduce GPU memory demand. However, small patch volumes capture less spatial context that leads to lower accuracy. As a pilot study, the purpose of this work is to propose a memory-efficient FCN which enables us to train the model on full size CT image directly without subvolume cropping, while maintaining the segmentation accuracy. We optimize our network from both architecture and implementation. With the development of computing hardware, such as graphics processing unit (GPU) and tensor processing unit (TPU), now deep learning applications is able to train networks with large datasets within acceptable time. Among these applications, semantic segmentation using fully convolutional network (FCN) also has gained a significant improvement against traditional image processing approaches in both computer vision and medical image processing fields. However, unlike general color images used in computer vision tasks, medical images have larger scales than color images such as 3D computed tomography (CT) images, micro CT images, and histopathological images. For training these medical images, the large demand of computing resource become a severe problem. In this paper, we present a memory-efficient FCN to tackle the high GPU memory demand challenge in organ segmentation problem from clinical CT images. The experimental results demonstrated that our GPU memory demand is about 40% of baseline architecture, parameter amount is about 30% of the baseline.

arXiv:2003.10658

CRNet: Cross-Reference Networks for Few-Shot Segmentation

Over the past few years, state-of-the-art image segmentation algorithms are based on deep convolutional neural networks. To render a deep network with the ability to understand a concept, humans need to collect a large amount of pixel-level annotated data to train the models, which is time-consuming and tedious. Recently, few-shot segmentation is proposed to solve this problem. Few-shot segmentation aims to learn a segmentation model that can be generalized to novel classes with only a few training images. In this paper, we propose a cross-reference network (CRNet) for few-shot segmentation. Unlike previous works which only predict the mask in the query image, our proposed model concurrently make predictions for both the support image and the query image. With a cross-reference mechanism, our network can better find the co-occurrent objects in the two images, thus helping the few-shot segmentation task. We also develop a mask refinement module to recurrently refine the prediction of the foreground regions. For the $k$-shot learning, we propose to finetune parts of networks to take advantage of multiple labeled support images. Experiments on the PASCAL VOC 2012 dataset show that our network achieves state-of-the-art performance.

arXiv:2003.10656공식 코드 · Papers with Code

Gen-LaneNet: A Generalized and Scalable Approach for 3D Lane Detection

We present a generalized and scalable method, called Gen-LaneNet, to detect 3D lanes from a single image. The method, inspired by the latest state-of-the-art 3D-LaneNet, is a unified framework solving image encoding, spatial transform of features and 3D lane prediction in a single network. However, we propose unique designs for Gen-LaneNet in two folds. First, we introduce a new geometry-guided lane anchor representation in a new coordinate frame and apply a specific geometric transformation to directly calculate real 3D lane points from the network output. We demonstrate that aligning the lane points with the underlying top-view features in the new coordinate frame is critical towards a generalized method in handling unfamiliar scenes. Second, we present a scalable two-stage framework that decouples the learning of image segmentation subnetwork and geometry encoding subnetwork. Compared to 3D-LaneNet, the proposed Gen-LaneNet drastically reduces the amount of 3D lane labels required to achieve a robust solution in real-world application. Moreover, we release a new synthetic dataset and its construction strategy to encourage the development and evaluation of 3D lane detection methods. In experiments, we conduct extensive ablation study to substantiate the proposed Gen-LaneNet significantly outperforms 3D-LaneNet in average precision(AP) and F-score.

arXiv:2003.09439공식 코드 · Papers with Code

ROAM: Random Layer Mixup for Semi-Supervised Learning in Medical Imaging

Medical image segmentation is one of the major challenges addressed by machine learning methods. Yet, deep learning methods profoundly depend on a large amount of annotated data, which is time-consuming and costly. Though, semi-supervised learning methods approach this problem by leveraging an abundant amount of unlabeled data along with a small amount of labeled data in the training process. Recently, MixUp regularizer has been successfully introduced to semi-supervised learning methods showing superior performance. MixUp augments the model with new data points through linear interpolation of the data at the input space. We argue that this option is limited. Instead, we propose ROAM, a RandOm lAyer Mixup, which encourages the network to be less confident for interpolated data points at randomly selected space. ROAM generates more data points that have never seen before, and hence it avoids over-fitting and enhances the generalization ability. We conduct extensive experiments to validate our method on three publicly available datasets on whole-brain image segmentation. ROAM achieves state-of-the-art (SOTA) results in fully supervised (89.5%) and semi-supervised (87.0%) settings with a relative improvement of up to 2.40% and 16.50%, respectively for the whole-brain segmentation.

arXiv:2003.10144

CF2-Net: Coarse-to-Fine Fusion Convolutional Network for Breast Ultrasound Image Segmentation

Breast ultrasound (BUS) image segmentation plays a crucial role in a computer-aided diagnosis system, which is regarded as a useful tool to help increase the accuracy of breast cancer diagnosis. Recently, many deep learning methods have been developed for segmentation of BUS image and show some advantages compared with conventional region-, model-, and traditional learning-based methods. However, previous deep learning methods typically use skip-connection to concatenate the encoder and decoder, which might not make full fusion of coarse-to-fine features from encoder and decoder. Since the structure and edge of lesion in BUS image are common blurred, these would make it difficult to learn the discriminant information of structure and edge, and reduce the performance. To this end, we propose and evaluate a coarse-to-fine fusion convolutional network (CF2-Net) based on a novel feature integration strategy (forming an 'E'-like type) for BUS image segmentation. To enhance contour and provide structural information, we concatenate a super-pixel image and the original image as the input of CF2-Net. Meanwhile, to highlight the differences in the lesion regions with variable sizes and relieve the imbalance issue, we further design a weighted-balanced loss function to train the CF2-Net effectively. The proposed CF2-Net was evaluated on an open dataset by using four-fold cross validation. The results of the experiment demonstrate that the CF2-Net obtains state-of-the-art performance when compared with other deep learning-based methods

arXiv:2003.08808

Deep Learning for Automatic Tracking of Tongue Surface in Real-time Ultrasound Videos, Landmarks instead of Contours

One usage of medical ultrasound imaging is to visualize and characterize human tongue shape and motion during a real-time speech to study healthy or impaired speech production. Due to the low-contrast characteristic and noisy nature of ultrasound images, it might require expertise for non-expert users to recognize tongue gestures in applications such as visual training of a second language. Moreover, quantitative analysis of tongue motion needs the tongue dorsum contour to be extracted, tracked, and visualized. Manual tongue contour extraction is a cumbersome, subjective, and error-prone task. Furthermore, it is not a feasible solution for real-time applications. The growth of deep learning has been vigorously exploited in various computer vision tasks, including ultrasound tongue contour tracking. In the current methods, the process of tongue contour extraction comprises two steps of image segmentation and post-processing. This paper presents a new novel approach of automatic and real-time tongue contour tracking using deep neural networks. In the proposed method, instead of the two-step procedure, landmarks of the tongue surface are tracked. This novel idea enables researchers in this filed to benefits from available previously annotated databases to achieve high accuracy results. Our experiment disclosed the outstanding performances of the proposed technique in terms of generalization, performance, and accuracy.

arXiv:2003.08740

Out-of-Distribution Detection in Multi-Label Datasets using Latent Space of $β$-VAE

Learning Enabled Components (LECs) are widely being used in a variety of perception based autonomy tasks like image segmentation, object detection, end-to-end driving, etc. These components are trained with large image datasets with multimodal factors like weather conditions, time-of-day, traffic-density, etc. The LECs learn from these factors during training, and while testing if there is variation in any of these factors, the components get confused resulting in low confidence predictions. The images with factors not seen during training is commonly referred to as Out-of-Distribution (OOD). For safe autonomy it is important to identify the OOD images, so that a suitable mitigation strategy can be performed. Classical one-class classifiers like SVM and SVDD are used to perform OOD detection. However, the multiple labels attached to the images in these datasets, restricts the direct application of these techniques. We address this problem using the latent space of the $\beta$-Variational Autoencoder ($\beta$-VAE). We use the fact that compact latent space generated by an appropriately selected $\beta$-VAE will encode the information about these factors in a few latent variables, and that can be used for computationally inexpensive detection. We evaluate our approach on the nuScenes dataset, and our results shows the latent space of $\beta$-VAE is sensitive to encode changes in the values of the generative factor.

arXiv:2003.08736

Real-Time High-Performance Semantic Image Segmentation of Urban Street Scenes

Deep Convolutional Neural Networks (DCNNs) have recently shown outstanding performance in semantic image segmentation. However, state-of-the-art DCNN-based semantic segmentation methods usually suffer from high computational complexity due to the use of complex network architectures. This greatly limits their applications in the real-world scenarios that require real-time processing. In this paper, we propose a real-time high-performance DCNN-based method for robust semantic segmentation of urban street scenes, which achieves a good trade-off between accuracy and speed. Specifically, a Lightweight Baseline Network with Atrous convolution and Attention (LBN-AA) is firstly used as our baseline network to efficiently obtain dense feature maps. Then, the Distinctive Atrous Spatial Pyramid Pooling (DASPP), which exploits the different sizes of pooling operations to encode the rich and distinctive semantic information, is developed to detect objects at multiple scales. Meanwhile, a Spatial detail-Preserving Network (SPN) with shallow convolutional layers is designed to generate high-resolution feature maps preserving the detailed spatial information. Finally, a simple but practical Feature Fusion Network (FFN) is used to effectively combine both shallow and deep features from the semantic branch (DASPP) and the spatial branch (SPN), respectively. Extensive experimental results show that the proposed method respectively achieves the accuracy of 73.6% and 68.0% mean Intersection over Union (mIoU) with the inference speed of 51.0 fps and 39.3 fps on the challenging Cityscapes and CamVid test datasets (by only using a single NVIDIA TITAN X card). This demonstrates that the proposed method offers excellent performance at the real-time speed for semantic segmentation of urban street scenes.

arXiv:2003.08462

Semi-supervised few-shot learning for medical image segmentation

Recent years have witnessed the great progress of deep neural networks on semantic segmentation, particularly in medical imaging. Nevertheless, training high-performing models require large amounts of pixel-level ground truth masks, which can be prohibitive to obtain in the medical domain. Furthermore, training such models in a low-data regime highly increases the risk of overfitting. Recent attempts to alleviate the need for large annotated datasets have developed training strategies under the few-shot learning paradigm, which addresses this shortcoming by learning a novel class from only a few labeled examples. In this context, a segmentation model is trained on episodes, which represent different segmentation problems, each of them trained with a very small labeled dataset. In this work, we propose a novel few-shot learning framework for semantic segmentation, where unlabeled images are also made available at each episode. To handle this new learning paradigm, we propose to include surrogate tasks that can leverage very powerful supervisory signals --derived from the data itself-- for semantic feature learning. We show that including unlabeled surrogate tasks in the episodic training leads to more powerful feature representations, which ultimately results in better generability to unseen tasks. We demonstrate the efficiency of our method in the task of skin lesion segmentation in two publicly available datasets. Furthermore, our approach is general and model-agnostic, which can be combined with different deep architectures.

arXiv:2003.07923

3D medical image segmentation with labeled and unlabeled data using autoencoders at the example of liver segmentation in CT images

Automatic segmentation of anatomical structures with convolutional neural networks (CNNs) constitutes a large portion of research in medical image analysis. The majority of CNN-based methods rely on an abundance of labeled data for proper training. Labeled medical data is often scarce, but unlabeled data is more widely available. This necessitates approaches that go beyond traditional supervised learning and leverage unlabeled data for segmentation tasks. This work investigates the potential of autoencoder-extracted features to improve segmentation with a CNN. Two strategies were considered. First, transfer learning where pretrained autoencoder features were used as initialization for the convolutional layers in the segmentation network. Second, multi-task learning where the tasks of segmentation and feature extraction, by means of input reconstruction, were learned and optimized simultaneously. A convolutional autoencoder was used to extract features from unlabeled data and a multi-scale, fully convolutional CNN was used to perform the target task of 3D liver segmentation in CT images. For both strategies, experiments were conducted with varying amounts of labeled and unlabeled training data. The proposed learning strategies improved results in $75\%$ of the experiments compared to training from scratch and increased the dice score by up to $0.040$ and $0.024$ for a ratio of unlabeled to labeled training data of about $32 : 1$ and $12.5 : 1$, respectively. The results indicate that both training strategies are more effective with a large ratio of unlabeled to labeled training data.

arXiv:2003.06529

Boundary Guidance Hierarchical Network for Real-Time Tongue Segmentation

Automated tongue image segmentation in tongue images is a challenging task for two reasons: 1) there are many pathological details on the tongue surface, which affect the extraction of the boundary; 2) the shapes of the tongues captured from various persons (with different diseases) are quite different. To deal with the challenge, a novel end-to-end Boundary Guidance Hierarchical Network (BGHNet) with a new hybrid loss is proposed in this paper. In the new approach, firstly Context Feature Encoder Module (CFEM) is built upon the bottomup pathway to confront with the shrinkage of the receptive field. Secondly, a novel hierarchical recurrent feature fusion module (HRFFM) is adopt to progressively and hierarchically refine object maps to recover image details by integrating local context information. Finally, the proposed hybrid loss in a four hierarchy-pixel, patch, map and boundary guides the network to effectively segment the tongue regions and accurate tongue boundaries. BGHNet is applied to a set of tongue images. The experimental results suggest that the proposed approach can achieve the latest tongue segmentation performance. And in the meantime, the lightweight network contains only 15.45M parameters and performs only 11.22GFLOPS.

arXiv:2003.07072

LT-Net: Label Transfer by Learning Reversible Voxel-wise Correspondence for One-shot Medical Image Segmentation

We introduce a one-shot segmentation method to alleviate the burden of manual annotation for medical images. The main idea is to treat one-shot segmentation as a classical atlas-based segmentation problem, where voxel-wise correspondence from the atlas to the unlabelled data is learned. Subsequently, segmentation label of the atlas can be transferred to the unlabelled data with the learned correspondence. However, since ground truth correspondence between images is usually unavailable, the learning system must be well-supervised to avoid mode collapse and convergence failure. To overcome this difficulty, we resort to the forward-backward consistency, which is widely used in correspondence problems, and additionally learn the backward correspondences from the warped atlases back to the original atlas. This cycle-correspondence learning design enables a variety of extra, cycle-consistency-based supervision signals to make the training process stable, while also boost the performance. We demonstrate the superiority of our method over both deep learning-based one-shot segmentation methods and a classical multi-atlas segmentation method via thorough experiments.

arXiv:2003.06778

A Simple Probabilistic Method for Deep Classification under Input-Dependent Label Noise

Datasets with noisy labels are a common occurrence in practical applications of classification methods. We propose a simple probabilistic method for training deep classifiers under input-dependent (heteroscedastic) label noise. We assume an underlying heteroscedastic generative process for noisy labels. To make gradient based training feasible we use a temperature parameterized softmax as a smooth approximation to the assumed generative process. We illustrate that the softmax temperature controls a bias-variance trade-off for the approximation. By tuning the softmax temperature, we improve accuracy, log-likelihood and calibration on both image classification benchmarks with controlled label noise as well as Imagenet-21k which has naturally occurring label noise. For image segmentation, our method increases the mean IoU on the PASCAL VOC and Cityscapes datasets by more than 1% over the state-of-the-art model.

arXiv:2003.05710

EDC3: Ensemble of Deep-Classifiers using Class-specific Copula functions to Improve Semantic Image Segmentation

In the literature, many fusion techniques are registered for the segmentation of images, but they primarily focus on observed output or belief score or probability score of the output classes. In the present work, we have utilized inter source statistical dependency among different classifiers for ensembling of different deep learning techniques for semantic segmentation of images. For this purpose, in the present work, a class-wise Copula-based ensembling method is newly proposed for solving the multi-class segmentation problem. Experimentally, it is observed that the performance has improved more for semantic image segmentation using the proposed class-specific Copula function than the traditionally used single Copula function for the problem. The performance is also compared with three state-of-the-art ensembling methods.

arXiv:2003.05056공식 코드 · Papers with Code

Multi-level Context Gating of Embedded Collective Knowledge for Medical Image Segmentation

Medical image segmentation has been very challenging due to the large variation of anatomy across different cases. Recent advances in deep learning frameworks have exhibited faster and more accurate performance in image segmentation. Among the existing networks, U-Net has been successfully applied on medical image segmentation. In this paper, we propose an extension of U-Net for medical image segmentation, in which we take full advantages of U-Net, Squeeze and Excitation (SE) block, bi-directional ConvLSTM (BConvLSTM), and the mechanism of dense convolutions. (I) We improve the segmentation performance by utilizing SE modules within the U-Net, with a minor effect on model complexity. These blocks adaptively recalibrate the channel-wise feature responses by utilizing a self-gating mechanism of the global information embedding of the feature maps. (II) To strengthen feature propagation and encourage feature reuse, we use densely connected convolutions in the last convolutional layer of the encoding path. (III) Instead of a simple concatenation in the skip connection of U-Net, we employ BConvLSTM in all levels of the network to combine the feature maps extracted from the corresponding encoding path and the previous decoding up-convolutional layer in a non-linear way. The proposed model is evaluated on six datasets DRIVE, ISIC 2017 and 2018, lung segmentation, $PH^2$, and cell nuclei segmentation, achieving state-of-the-art performance.

arXiv:2003.04949공식 코드 · Papers with Code

LC-GAN: Image-to-image Translation Based on Generative Adversarial Network for Endoscopic Images

Intelligent vision is appealing in computer-assisted and robotic surgeries. Vision-based analysis with deep learning usually requires large labeled datasets, but manual data labeling is expensive and time-consuming in medical problems. We investigate a novel cross-domain strategy to reduce the need for manual data labeling by proposing an image-to-image translation model live-cadaver GAN (LC-GAN) based on generative adversarial networks (GANs). We consider a situation when a labeled cadaveric surgery dataset is available while the task is instrument segmentation on an unlabeled live surgery dataset. We train LC-GAN to learn the mappings between the cadaveric and live images. For live image segmentation, we first translate the live images to fake-cadaveric images with LC-GAN and then perform segmentation on the fake-cadaveric images with models trained on the real cadaveric dataset. The proposed method fully makes use of the labeled cadaveric dataset for live image segmentation without the need to label the live dataset. LC-GAN has two generators with different architectures that leverage the deep feature representation learned from the cadaveric image based segmentation task. Moreover, we propose the structural similarity loss and segmentation consistency loss to improve the semantic consistency during translation. Our model achieves better image-to-image translation and leads to improved segmentation performance in the proposed cross-domain segmentation task.

arXiv:2003.04797

Dam Burst: A region-merging-based image segmentation method

Until now, all single level segmentation algorithms except CNN-based ones lead to over segmentation. And CNN-based segmentation algorithms have their own problems. To avoid over segmentation, multiple thresholds of criteria are adopted in region merging process to produce hierarchical segmentation results. However, there still has extreme over segmentation in the low level of the hierarchy, and outstanding tiny objects are merged to their large adjacencies in the high level of the hierarchy. This paper proposes a region-merging-based image segmentation method that we call it Dam Burst. As a single level segmentation algorithm, this method avoids over segmentation and retains details by the same time. It is named because of that it simulates a flooding from underground destroys dams between water-pools. We treat edge detection results as strengthening structure of a dam if it is on the dam. To simulate a flooding from underground, regions are merged by ascending order of the average gra-dient inside the region.

arXiv:2003.04377공식 코드 · Papers with Code

Automatic segmentation of spinal multiple sclerosis lesions: How to generalize across MRI contrasts?

Despite recent improvements in medical image segmentation, the ability to generalize across imaging contrasts remains an open issue. To tackle this challenge, we implement Feature-wise Linear Modulation (FiLM) to leverage physics knowledge within the segmentation model and learn the characteristics of each contrast. Interestingly, a well-optimised U-Net reached the same performance as our FiLMed-Unet on a multi-contrast dataset (0.72 of Dice score), which suggests that there is a bottleneck in spinal MS lesion segmentation different from the generalization across varying contrasts. This bottleneck likely stems from inter-rater variability, which is estimated at 0.61 of Dice score in our dataset.

arXiv:2003.03862

Improving Training on Noisy Stuctured Labels

Fine-grained annotations---e.g. dense image labels, image segmentation and text tagging---are useful in many ML applications but they are labor-intensive to generate. Moreover there are often systematic, structured errors in these fine-grained annotations. For example, a car might be entirely unannotated in the image, or the boundary between a car and street might only be coarsely annotated. Standard ML training on data with such structured errors produces models with biases and poor performance. In this work, we propose a novel framework of Error-Correcting Networks (ECN) to address the challenge of learning in the presence structured error in fine-grained annotations. Given a large noisy dataset with commonly occurring structured errors, and a much smaller dataset with more accurate annotations, ECN is able to substantially improve the prediction of fine-grained annotations compared to standard approaches for training on noisy data. It does so by learning to leverage the structures in the annotations and in the noisy labels. Systematic experiments on image segmentation and text tagging demonstrate the strong performance of ECN in improving training on noisy structured labels.

arXiv:2003.03744

A Multi-scale CNN-CRF Framework for Environmental Microorganism Image Segmentation

In order to assist researchers to identify Environmental Microorganisms (EMs) effectively, a Multi-scale CNN-CRF (MSCC) framework for the EM image segmentation is proposed in this paper. There are two parts in this framework: The first is a novel pixel-level segmentation approach, using a newly introduced Convolutional Neural Network (CNN), namely "mU-Net-B3", with a dense Conditional Random Field (CRF) post-processing. The second is a VGG-16 based patch-level segmentation method with a novel "buffer" strategy, which further improves the segmentation quality of the details of the EMs. In the experiment, compared with the state-of-the-art methods on 420 EM images, the proposed MSCC method reduces the memory requirement from 355 MB to 103 MB, improves the overall evaluation indexes (Dice, Jaccard, Recall, Accuracy) from 85.24%, 77.42%, 82.27% and 96.76% to 87.13%, 79.74%, 87.12% and 96.91% respectively, and reduces the volume overlap error from 22.58% to 20.26%. Therefore, the MSCC method shows a big potential in the EM segmentation field.