The convolutional neural network (CNN) and other neural networks (NNs) provide promising tools for robotized characterization of tumor cells. However, the tumor growth areas in ultrasound images are normally obscure, ...
详细信息
The convolutional neural network (CNN) and other neural networks (NNs) provide promising tools for robotized characterization of tumor cells. However, the tumor growth areas in ultrasound images are normally obscure, with uncertain edges. It is not acceptable to prepare ultrasound images straightforwardly with the CNN. To solve the problem, this paper puts forward a faster region-convolutional neural network (R-CNN) to identify tumor cells with the aid of auto encoders Taking two fully-connected layers with dropout and ReLU enactments as the base, the proposed faster R-CNN adopts 3D convolutional and max pooling layers, enabling the user to extract features from potential tumor growth areas. In addition, the thin and deep layers of the network were connected to facilitate the identification of blurry or small tumor growth areas. Experimental results show that the proposed faster R-CNN with auto encoders outperformed traditional data mining and artificial intelligence (AI) methods in prediction accuracy of tumor cells.
Accurate analysis of meibomian gland morphology based on meibography images is of great importance for the diagnosis of dry eye disease. However, it is still a difficult task due to the time-consuming and variability ...
详细信息
Saliency-driven image and video coding for humans has gained importance in the recent past. In this paper, we pro-pose such a saliency-driven coding framework for the video coding for machines task using the latest vi...
详细信息
ISBN:
(纸本)9781728176055;9781728176062
Saliency-driven image and video coding for humans has gained importance in the recent past. In this paper, we pro-pose such a saliency-driven coding framework for the video coding for machines task using the latest video coding standard Versatile Video Coding (VVC). To determine the salient regions before encoding, we employ the real-time-capable object detection network You Only Look Once (YOLO) in combination with a novel decision criterion. To measure the coding quality for a machine, the state-of-the-art object segmentation network Mask R-CNN was applied to the decoded frame. From extensive simulations we find that, compared to the reference VVC with a constant quality, up to 29 % of bitrate can be saved with the same detection accuracy at the decoder side by applying the proposed saliency-driven framework. Besides, we compare YOLO against other, more traditional saliency detection methods.
To solve the problem of underwater proud object classification using high-resolution sonar image under small sample situation,a classification method using deep neural network is ***,statistical characteristics of aco...
详细信息
To solve the problem of underwater proud object classification using high-resolution sonar image under small sample situation,a classification method using deep neural network is ***,statistical characteristics of acoustic shadow regions are modeled using Gaussian mixture model and acoustic shadow is *** and simulated dataset are constructed on this ***,simulated dataset is input into convolutional neural network for training,and the feature extraction part is retained,which is used to extract feature of trial *** classification part is reconstructed and trained by feature vectors of trial *** experimental results show that the average classification accuracy of the proposed method is 88.24%,which is 8.67%,20.47%,19.78%,11.59%,9.01%,11.58% higher than that of other six methods *** verifies that the proposed method achieves better performance on underwater proud object classification *** learning curve converges to 96.25%,which is 5.14% higher than validation curve,indicating that the over-fitting problem is alleviated to some *** convolutional neural network is applied in a fusion classifier,which also combines output of logistic classifier,support vector machine,and finally obtains a fusion *** classification accuracy is up to 93.33%,indicating that fusion classifier improves robustness and classification performance of algorithm *** proposed method combines deep learning and transfer learning,which not only utilizes powerful image classification ability of convolutional neural network,but also avoids serious over-fitting problem caused by limited dataset.
While deep neural networks (DNNs) have achieved state-of-the-art results in many fields, they are typically over-parameterized. Parameter redundancy, in turn, leads to inefficiency. Sparse signal recovery (SSR) techni...
详细信息
ISBN:
(数字)9781509066315
ISBN:
(纸本)9781509066315;9781509066322
While deep neural networks (DNNs) have achieved state-of-the-art results in many fields, they are typically over-parameterized. Parameter redundancy, in turn, leads to inefficiency. Sparse signal recovery (SSR) techniques, on the other hand, find compact solutions to over-complete linear problems. Therefore, a logical step is to draw the connection between SSR and DNNs. In this paper, we explore the application of iterative reweighting methods popular in SSR to learning efficient DNNs. By efficient, we mean sparse networks that require less computation and storage than the original, dense network. We propose a reweighting framework to learn sparse connections within a given architecture without biasing the optimization process, by utilizing the affine scaling transformation strategy. The resulting algorithm, referred to as Sparsity-promoting stochastic Gradient Descent (SSGD), has simple gradient-based updates which can be easily implemented in existing deep learning libraries. We demonstrate the sparsification ability of SSGD on image classification tasks and show that it outperforms existing methods on the MNIST and CIFAR-10 datasets.
The proceedings contain 83 papers. The special focus in this conference is on Machine Learning, imageprocessing, Network Security and Data Sciences. The topics include: An Empirical Study to Predict Myocardial Infarc...
ISBN:
(纸本)9789811563171
The proceedings contain 83 papers. The special focus in this conference is on Machine Learning, imageprocessing, Network Security and Data Sciences. The topics include: An Empirical Study to Predict Myocardial Infarction Using K-Means and Hierarchical Clustering;a Robust Technique for End Point Detection Under Practical Environment;an Explainable Machine Learning Approach for Definition Extraction;Steps of Pre-processing for English to Mizo SMT System;efficient Human Feature Recognition Process Using Sclera;optimization of Local Ordering Technique for Nearest Neighbour Circuits;attention-Based English to Mizo neural Machine Translation;in Depth Analysis of Lung Disease Prediction Using Machine Learning Algorithms;improve the Accuracy of Heart Disease Predictions Using Machine Learning and Feature Selection Techniques;Evaluation of Multiplier-Less DCT Transform Using In-Exact Computing;convolutional neural Network Based Sound Recognition methods for Detecting Presence of Amateur Drones in Unauthorized Zones;comparison of Different Decision Tree Algorithms for Predicting the Heart Disease;dynamic Speech Trajectory Based Parameters for Low Resource Languages;identification and Prediction of Alzheimer Based on Biomarkers Using ‘Machine Learning’;solving Quadratic Assignment Problem Using Crow Search Algorithm in Accelerated Systems;speech signal Analysis for Language Identification Using Tensors;Effective Removal of Baseline Wander from ECG signals: A Comparative Study;Face Recognition Based on Human Sketches Using Fuzzy Minimal Structure Oscillation in the SIFT Domain;an Ensemble Model for Predicting Passenger Demand Using Taxi Data Set;a Novel Approach to Synthesize Hinglish Text to English Text;legal Amount Recognition in Bank Cheques Using Capsule Networks;comparative Analysis of neural Models for Abstractive Text Summarization.
Fluorescence lifetime imaging microscopy (FLIM) systems are limited by their slow processing speed, low signalto- noise ratio (SNR), and expensive and challenging hardware setups. In this work, we demonstrate applying...
详细信息
Although color images are easily interfered by glass, depth images captured by infrared sensors are robust to reflection. In this letter, we propose multi-modal reflection removal using convolutional neural networks (...
详细信息
Although color images are easily interfered by glass, depth images captured by infrared sensors are robust to reflection. In this letter, we propose multi-modal reflection removal using convolutional neural networks (CNNs). We build a multi-modal CNN for reflection removal to separate transmission from reflection using depth information. The proposed network consists of two sub-networks: image restoration and depth adaptation. image restoration sub-network (IRN) recovers transmission layer from the input image with reflection, whereas depth adaptation subnetwork (DAN) guides reflection removal of the IRN. Moreover, to extract image details for reflection removal, we present a multi-scale loss function that penalizes non-similarity for multi-scale outputs. Experimental results demonstrate that the proposed method is robust to dominant reflections and outperforms state-of-the-art methods in terms of both peak signal-to-noise ratio (PSNR) and structural similarity.
In today's data- and computation-driven society, day-to-day life depends on devices such as smartphones, laptops, smart watches, and biosensors/image sensors connected to computational engines. The computationall...
详细信息
In today's data- and computation-driven society, day-to-day life depends on devices such as smartphones, laptops, smart watches, and biosensors/image sensors connected to computational engines. The computationally intensive applications that run on these devices incur high levels of chip power dissipation, and must operate under stringent power constraints due to thermal or battery life limitations. On future hardware platforms, a large fraction of computation power will be spent on error-tolerant multimedia applications such as signalprocessing tasks (on audio, video, or images) and artificial intelligence (AI) algorithms for recognizing voice and image data. For such error-tolerant applications, approximate computation has emerged as a new paradigm that provides a pragmatic approach for trading off energy/power for computational accuracy. A powerful method for implementing approximate computing is by performing logic-level or architecture-level hardware modifications. The effectiveness of an approximate system depends on identifying potential modes of approximation, accurate modeling of injected error as a function of the approximation, and optimization of the system to maximize energy savings for user-defined quality constraints. However, current approaches to approximate computation involve ad hoc trial-and-error based methods that do not consider the effect of approximations on system-level quality metrics. Additionally, prior methods for approximate computation have provided little or no scope for modulating the design based on user- and application-specific error budgets. HASH(0x4210e28) This thesis proposes adaptive frameworks for energy-efficient approximate computing, leveraging the target application characteristics, system architecture, and input information to build fast, power-efficient approximate circuits under a user-defined error budget. The work is focused on two well-established, widely-used, and computationally intensive applications: multim
暂无评论