Despeckling CNN with Ensembles of Classical Outputs
01 Aug 2018-pp 3802-3807
TL;DR: A convolutional neural network is developed which learns to remove speckle from US images using the outputs of these classical approaches and is able to outperform the state-of-the-art despeckling approaches and produces the outputs even better than the ensembles for certain images.
Abstract: Ultrasound (US) image despeckling is a problem of high clinical importance. Machine learning solutions to the problem are considered impractical due to the unavailability of speckle-free US image dataset. On the other hand, the classical approaches, which are able to provide the desired outputs, have limitations like input dependent parameter tuning. In this work, a convolutional neural network (CNN) is developed which learns to remove speckle from US images using the outputs of these classical approaches. It is observed that the existing approaches can be combined in a complementary manner to generate an output better than their individual outputs. Thus, the CNN is trained using the individual outputs as well as the output ensembles. It eliminates the cumbersome process of parameter tuning required by the existing approaches for every new input. Further, the proposed CNN is able to outperform the state-of-the-art despeckling approaches and produces the outputs even better than the ensembles for certain images.
Citations
More filters
TL;DR: This paper describes, in detail, 27 techniques that mainly focus on the smoothing or elimination of speckle noise in medical ultrasound images, and describes recent techniques in the field of machine learning focused on deep learning, which are not yet well known but greatly relevant.
Abstract: In recent years, many studies have examined filters for eliminating or reducing speckle noise, which is inherent to ultrasound images, in order to improve the metrological evaluation of their biomedical applications. In the case of medical ultrasound images, said noise can produce uncertainty in the diagnosis because details, such as limits and edges, should be preserved. Most algorithms can eliminate speckle noise, but they do not consider the conservation of these details. This paper describes, in detail, 27 techniques that mainly focus on the smoothing or elimination of speckle noise in medical ultrasound images. The aim of this study is to highlight the importance of improving said smoothing and elimination, which are directly related to several processes (such as the detection of regions of interest) described in other articles examined in this study. Furthermore, the description of this collection of techniques facilitates the implementation of evaluations and research with a more specific scope. This study initially covers several classical methods, such as spatial filtering, diffusion filtering, and wavelet filtering. Subsequently, it describes recent techniques in the field of machine learning focused on deep learning, which are not yet well known but greatly relevant, along with some modern and hybrid models in the field of speckle-noise filtering. Finally, five Full-Reference (FR) distortion metrics, common in filter evaluation processes, are detailed along with a compensation methodology between FR and Non-Reference (NR) metrics, which can generate greater certainty in the classification of the filters by considering the information of their behavior in terms of perceptual quality provided by NR metrics.
28 citations
Cites result from "Despeckling CNN with Ensembles of C..."
...In contrast, other authors [87] adopted a Deep Learning approach....
[...]
Posted Content•
TL;DR: This paper aims at helping researchers and practitioners to better understand the application of ML techniques to RSP-related problems by providing a comprehensive, structured and reasoned literature overview of ML-based RSP techniques.
Abstract: Modern radar systems have high requirements in terms of accuracy, robustness and real-time capability when operating on increasingly complex electromagnetic environments. Traditional radar signal processing (RSP) methods have shown some limitations when meeting such requirements, particularly in matters of target classification. With the rapid development of machine learning (ML), especially deep learning, radar researchers have started integrating these new methods when solving RSP-related problems. This paper aims at helping researchers and practitioners to better understand the application of ML techniques to RSP-related problems by providing a comprehensive, structured and reasoned literature overview of ML-based RSP techniques. This work is amply introduced by providing general elements of ML-based RSP and by stating the motivations behind them. The main applications of ML-based RSP are then analysed and structured based on the application field. This paper then concludes with a series of open questions and proposed research directions, in order to indicate current gaps and potential future solutions and trends.
22 citations
Cites methods from "Despeckling CNN with Ensembles of C..."
...In addition, combined with ensemble learning method, the authors proposed a despecking CNN architecture in [272]....
[...]
TL;DR: The experiment shows that the compression of 86.2% can be achieved using the threshold of two intensity levels and the compressed image can be reconstructed with the PSNR of 45.87 dB.
Abstract: A superpixel based on-chip compression is proposed in this paper. Pixels are compared in spatial domain and the pixels with similar characteristics are grouped to form the superpixels. Only one pixel corresponding to each superpixel is read to achieve the compression. The on-chip compression circuit is designed and simulated in UMC 180 nm CMOS technology. For 70% compression, the proposed design results in about 33% power saving. The reconstruction of the compressed image is performed off-chip using bilinear interpolation. Further, two enhancement approaches are developed to improve the output image quality. The first approach is based on wavelet decomposition whereas the second approach uses a deep convolutional neural network. The proposed reconstruction technique takes two orders of magnitude lesser time as compared to the state-of-the-art technique. On an average, it results in peak signal to noise ratio (PSNR) and structural similarity index measure values of 30.999 and 0.9088 dB, respectively, for 70% compression in natural images. On the other hand, the best values observed from the existing approaches for the two metrics are 28.634 and 0.8115 dB, respectively. Further, the proposed technique is found useful for thermal image compression and reconstruction. The experiment shows that the compression of 86.2% can be achieved using the threshold of two intensity levels and the compressed image can be reconstructed with the PSNR of 45.87 dB.
8 citations
Cites background from "Despeckling CNN with Ensembles of C..."
...The networks used in [33] and [34] have a limited receptive field due to the absence of pooling layers....
[...]
...FCNNs have shown recent success in application like noise reduction [33], [34]....
[...]
...targeted in [33] and [34], the discontinuities occurring in the interpolated images are generated due to long streaks of the...
[...]
TL;DR: The results show that the efficiency of the filtration strongly depends on the specific wavelet system setting, type of ultrasound data, and the noise present, which may provide a useful guideline for researchers, software developers, and clinical professionals to obtain high quality images.
Abstract: Wavelet transform (WT) is a commonly used method for noise suppression and feature extraction from biomedical images. The selection of WT system settings significantly affects the efficiency of denoising procedure. This comparative study analyzed the efficacy of the proposed WT system on real 292 ultrasound images from several areas of interest. The study investigates the performance of the system for different scaling functions of two basic wavelet bases, Daubechies and Symlets, and their efficiency on images artificially corrupted by three kinds of noise. To evaluate our extensive analysis, we used objective metrics, namely structural similarity index (SSIM), correlation coefficient, mean squared error (MSE), peak signal-to-noise ratio (PSNR) and universal image quality index (Q-index). Moreover, this study includes clinical insights on selected filtration outcomes provided by clinical experts. The results show that the efficiency of the filtration strongly depends on the specific wavelet system setting, type of ultrasound data, and the noise present. The findings presented may provide a useful guideline for researchers, software developers, and clinical professionals to obtain high quality images.
1 citations
References
More filters
27 Jun 2016
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.
123,388 citations
Proceedings Article•
01 Jan 2015TL;DR: This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.
Abstract: We introduce Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. The method is straightforward to implement, is computationally efficient, has little memory requirements, is invariant to diagonal rescaling of the gradients, and is well suited for problems that are large in terms of data and/or parameters. The method is also appropriate for non-stationary objectives and problems with very noisy and/or sparse gradients. The hyper-parameters have intuitive interpretations and typically require little tuning. Some connections to related algorithms, on which Adam was inspired, are discussed. We also analyze the theoretical convergence properties of the algorithm and provide a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework. Empirical results demonstrate that Adam works well in practice and compares favorably to other stochastic optimization methods. Finally, we discuss AdaMax, a variant of Adam based on the infinity norm.
111,197 citations
"Despeckling CNN with Ensembles of C..." refers methods in this paper
...Adam optimizer [27] with an initial learning rate of 10−4 and decay factor of 10−7 are used to optimize the parameters....
[...]
Proceedings Article•
03 Dec 2012TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Abstract: We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overriding in the fully-connected layers we employed a recently-developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.
73,978 citations
TL;DR: In this article, a structural similarity index is proposed for image quality assessment based on the degradation of structural information, which can be applied to both subjective ratings and objective methods on a database of images compressed with JPEG and JPEG2000.
Abstract: Objective methods for assessing perceptual image quality traditionally attempted to quantify the visibility of errors (differences) between a distorted image and a reference image using a variety of known properties of the human visual system. Under the assumption that human visual perception is highly adapted for extracting structural information from a scene, we introduce an alternative complementary framework for quality assessment based on the degradation of structural information. As a specific example of this concept, we develop a structural similarity index and demonstrate its promise through a set of intuitive examples, as well as comparison to both subjective ratings and state-of-the-art objective methods on a database of images compressed with JPEG and JPEG2000. A MATLAB implementation of the proposed algorithm is available online at http://www.cns.nyu.edu//spl sim/lcv/ssim/.
40,609 citations
[...]
TL;DR: This work presents a conceptually simple, flexible, and general framework for object instance segmentation, which extends Faster R-CNN by adding a branch for predicting an object mask in parallel with the existing branch for bounding box recognition.
Abstract: We present a conceptually simple, flexible, and general framework for object instance segmentation. Our approach efficiently detects objects in an image while simultaneously generating a high-quality segmentation mask for each instance. The method, called Mask R-CNN, extends Faster R-CNN by adding a branch for predicting an object mask in parallel with the existing branch for bounding box recognition. Mask R-CNN is simple to train and adds only a small overhead to Faster R-CNN, running at 5 fps. Moreover, Mask R-CNN is easy to generalize to other tasks, e.g., allowing us to estimate human poses in the same framework. We show top results in all three tracks of the COCO suite of challenges, including instance segmentation, bounding-box object detection, and person keypoint detection. Without tricks, Mask R-CNN outperforms all existing, single-model entries on every task, including the COCO 2016 challenge winners. We hope our simple and effective approach will serve as a solid baseline and help ease future research in instance-level recognition. Code will be made available.
14,299 citations
"Despeckling CNN with Ensembles of C..." refers methods in this paper
...which have been successfully applied to the computer vision application ranging from object recognition [21] and classification [22], [23] to noise reduction [17]....
[...]