scholarly journals Interpreting and Comparing Convolutional Neural Networks: A Quantitative Approach

2021 ◽  
Author(s):  
Mohammad Mohaiminul Islam ◽  
Zahid Hassan Tushar
Keyword(s):  
Neural Network ◽  
Region Of Interest ◽  
Input Image ◽  
Black Box ◽  
Discriminative Ability ◽  
Highly Active ◽  
Semantic Concepts ◽  
Gradient Based ◽  
Visual Concepts ◽  
Hidden Neurons

A convolutional neural network (CNN) is sometimes understood as a black box in the sense that while it can approximate any function, studying its structure will not give us any insights into the nature of the function being approximated. In other terms, the discriminative ability does not reveal much about the latent representation of a network. This research aims to establish a framework for interpreting the CNNs by profiling them in terms of interpretable visual concepts and verifying them by means of Integrated Gradient. We also ask the question, "Do different input classes have a relationship or are they unrelated?" For instance, could there be an overlapping set of highly active neurons to identify different classes? Could there be a set of neurons that are useful for one input class whereas misleading for a different one? Intuition answers these questions positively, implying the existence of a structured set of neurons inclined to a particular class. Knowing this structure has significant values; it provides a principled way for identifying redundancies across the classes. Here the interpretability profiling has been done by evaluating the correspondence between individual hidden neurons and a set of human-understandable visual semantic concepts. We also propose an integrated gradient-based class-specific relevance mapping approach that takes the spatial position of the region of interest in the input image. Our relevance score verifies the interpretability scores in terms of neurons tuned to a particular concept/class. Further, we perform network ablation and measure the performance of the network based on our approach.

scholarly journals Interpreting and Comparing Convolutional Neural Networks: A Quantitative Approach

2021 ◽  
Author(s):  
Mohammad Mohaiminul Islam ◽  
Zahid Hassan Tushar
Keyword(s):  
Neural Network ◽  
Region Of Interest ◽  
Input Image ◽  
Black Box ◽  
Discriminative Ability ◽  
Highly Active ◽  
Semantic Concepts ◽  
Gradient Based ◽  
Visual Concepts ◽  
Hidden Neurons

A convolutional neural network (CNN) is sometimes understood as a black box in the sense that while it can approximate any function, studying its structure will not give us any insights into the nature of the function being approximated. In other terms, the discriminative ability does not reveal much about the latent representation of a network. This research aims to establish a framework for interpreting the CNNs by profiling them in terms of interpretable visual concepts and verifying them by means of Integrated Gradient. We also ask the question, "Do different input classes have a relationship or are they unrelated?" For instance, could there be an overlapping set of highly active neurons to identify different classes? Could there be a set of neurons that are useful for one input class whereas misleading for a different one? Intuition answers these questions positively, implying the existence of a structured set of neurons inclined to a particular class. Knowing this structure has significant values; it provides a principled way for identifying redundancies across the classes. Here the interpretability profiling has been done by evaluating the correspondence between individual hidden neurons and a set of human-understandable visual semantic concepts. We also propose an integrated gradient-based class-specific relevance mapping approach that takes the spatial position of the region of interest in the input image. Our relevance score verifies the interpretability scores in terms of neurons tuned to a particular concept/class. Further, we perform network ablation and measure the performance of the network based on our approach.

scholarly journals Pose Invariant Palm Vein Identification System using Convolutional Neural Network

2018 ◽  
Vol 15 (4) ◽  
pp. 502-509 ◽  
Author(s):  
Baghdad Science Journal
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Data Augmentation ◽  
Region Of Interest ◽  
Input Image ◽  
Identification System ◽  
Plane Rotation ◽  
Imaging Device ◽  
Out Of Plane ◽  
Palm Vein

Palm vein recognition is a one of the most efficient biometric technologies, each individual can be identified through its veins unique characteristics, palm vein acquisition techniques is either contact based or contactless based, as the individual's hand contact or not the peg of the palm imaging device, the needs a contactless palm vein system in modern applications rise tow problems, the pose variations (rotation, scaling and translation transformations) since the imaging device cannot aligned correctly with the surface of the palm, and a delay of matching process especially for large systems, trying to solve these problems. This paper proposed a pose invariant identification system for contactless palm vein which include three main steps, at first data augmentation is done by making multiple copies of the input image then perform out-of-plane rotation on them around all the X,Y and Z axes. Then a new fast extract Region of Interest (ROI) algorithm is proposed for cropping palm region. Finally, features are extracted and classified by specific structure of Convolutional Neural Network (CNN). The system is tested on two public multispectral palm vein databases (PolyU and CASIA); furthermore, synthetic datasets are derived from these mentioned databases, to simulate the hand out-of-plane rotation in random angels within range from -20° to +20° degrees. To study several situations of pose invariant, twelve experiments are performed on all datasets, highest accuracy achieved is 99.73% ∓ 0.27 on PolyU datasets and 98 % ∓ 1 on CASIA datasets, with very fast identification process, about 0.01 second for identifying an individual, which proves system efficiency in contactless palm vein problems.

scholarly journals A Hybrid Compression Method for Medical Images Based on Region of Interest Using Artificial Neural Networks

2021 ◽  
Vol 2021 ◽  
pp. 1-9
Author(s):  
Ali Ibrahim Khaleel ◽  
Nik Adilah Hanin Zahri ◽  
Muhammad Imran Ahmad
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
State Of The Art ◽  
Medical Images ◽  
Region Of Interest ◽  
Input Image ◽  
Compression Method ◽  
Medical Diagnoses ◽  
The Neural Network ◽  
Similar Images

The number of medical images being stored and communicated daily is rapidly increasing, according to the need for these images in medical diagnoses. Hence, the storage space and bandwidths required to store and communicate these images are exponentially increasing, which has brought attention toward compressing these images. In this study, a new compression method is proposed for medical images based on convolutional neural networks. The proposed neural network consists of two main stages: a segmentation stage and an autoencoder. The segmentation stage is used to recognize the Region of Interest (ROI) in the image and provide it to the autoencoder stage, so more emphasis on the information of the ROI is applied. The autoencoder part of the neural network contains a bottleneck layer that has one-eighth of the dimensions of the input image. The values in this layer are used to represent the image, while the following layers are used to decompress the images, after training the neural network. The proposed method is evaluated using the CLEF MED 2009 dataset, where the evaluation results show that the method has significantly better performance, compared to the existing state-of-the-art methods, by providing more visually similar images using less data.

scholarly journals CoCoX: Generating Conceptual and Counterfactual Explanations via Fault-Lines

2020 ◽  
Vol 34 (03) ◽  
pp. 2594-2601
Author(s):  
Arjun Akula ◽  
Shuai Wang ◽  
Song-Chun Zhu
Keyword(s):  
Neural Network ◽  
State Of The Art ◽  
Input Image ◽  
Classification Model ◽  
Learning Models ◽  
Fault Line ◽  
Semantic Level ◽  
Explainable Ai ◽  
Fault Lines ◽  
Classification Category

We present CoCoX (short for Conceptual and Counterfactual Explanations), a model for explaining decisions made by a deep convolutional neural network (CNN). In Cognitive Psychology, the factors (or semantic-level features) that humans zoom in on when they imagine an alternative to a model prediction are often referred to as fault-lines. Motivated by this, our CoCoX model explains decisions made by a CNN using fault-lines. Specifically, given an input image I for which a CNN classification model M predicts class cpred, our fault-line based explanation identifies the minimal semantic-level features (e.g., stripes on zebra, pointed ears of dog), referred to as explainable concepts, that need to be added to or deleted from I in order to alter the classification category of I by M to another specified class calt. We argue that, due to the conceptual and counterfactual nature of fault-lines, our CoCoX explanations are practical and more natural for both expert and non-expert users to understand the internal workings of complex deep learning models. Extensive quantitative and qualitative experiments verify our hypotheses, showing that CoCoX significantly outperforms the state-of-the-art explainable AI models. Our implementation is available at https://github.com/arjunakula/CoCoX

Sparse-FCM and Deep Convolutional Neural Network for the segmentation and classification of acute lymphoblastic leukaemia

2020 ◽  
Vol 65 (6) ◽  
pp. 759-773
Author(s):  
Segu Praveena ◽  
Sohan Pal Singh
Keyword(s):  
Neural Network ◽  
Acute Lymphoblastic Leukaemia ◽  
Convolutional Neural Network ◽  
Optimization Algorithm ◽  
Lymphoblastic Leukaemia ◽  
Input Image ◽  
Deep Convolutional Neural Network ◽  
Grey Wolf ◽  
Deep Cnn

AbstractLeukaemia detection and diagnosis in advance is the trending topic in the medical applications for reducing the death toll of patients with acute lymphoblastic leukaemia (ALL). For the detection of ALL, it is essential to analyse the white blood cells (WBCs) for which the blood smear images are employed. This paper proposes a new technique for the segmentation and classification of the acute lymphoblastic leukaemia. The proposed method of automatic leukaemia detection is based on the Deep Convolutional Neural Network (Deep CNN) that is trained using an optimization algorithm, named Grey wolf-based Jaya Optimization Algorithm (GreyJOA), which is developed using the Grey Wolf Optimizer (GWO) and Jaya Optimization Algorithm (JOA) that improves the global convergence. Initially, the input image is applied to pre-processing and the segmentation is performed using the Sparse Fuzzy C-Means (Sparse FCM) clustering algorithm. Then, the features, such as Local Directional Patterns (LDP) and colour histogram-based features, are extracted from the segments of the pre-processed input image. Finally, the extracted features are applied to the Deep CNN for the classification. The experimentation evaluation of the method using the images of the ALL IDB2 database reveals that the proposed method acquired a maximal accuracy, sensitivity, and specificity of 0.9350, 0.9528, and 0.9389, respectively.

Gradient-based multi-focus image fusion method using convolution neural network

2021 ◽  
Vol 92 ◽  
pp. 107174
Author(s):  
Yang Zhou ◽  
Xiaomin Yang ◽  
Rongzhu Zhang ◽  
Kai Liu ◽  
Marco Anisetti ◽  
...  
Keyword(s):  
Neural Network ◽  
Image Fusion ◽  
Convolution Neural Network ◽  
Fusion Method ◽  
Gradient Based ◽  
Focus Image ◽  
Image Fusion Method

scholarly journals Opening the black box of neural networks: methods for interpreting neural network models in clinical applications

2018 ◽  
Vol 6 (11) ◽  
pp. 216-216 ◽  
Author(s):  
Zhongheng Zhang ◽  
◽  
Marcus W. Beck ◽  
David A. Winkler ◽  
Bin Huang ◽  
...  
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Network Models ◽  
Black Box ◽  
Clinical Applications ◽  
Neural Network Models

Prediction of Surface Roughness in Magneto Rheological Abrasive Flow Finishing Process by Artificial Neural Networks and Regression Analysis

2015 ◽  
Vol 766-767 ◽  
pp. 1076-1084
Author(s):  
S. Kathiresan ◽  
K. Hariharan ◽  
B. Mohan
Keyword(s):  
Neural Network ◽  
Surface Roughness ◽  
Regression Analysis ◽  
Back Propagation ◽  
Hydraulic Pressure ◽  
Artificial Neural ◽  
Artificial Neural Network Ann ◽  
Magneto Rheological ◽  
Hidden Neurons ◽  
Abrasive Flow Finishing

In this study, to predict the surface roughness of stainless steel-304 in Magneto rheological Abrasive flow finishing (MRAFF) process, an artificial neural network (ANN) and regression models have been developed. In this models, the parameters such as hydraulic pressure, current to the electromagnet and number of cycles were taken as variables of the model.Taguchi’s technique has been used for designing the experiments in order to observe the different values of surface roughness . A neural network with feed forward with the help of back propagation was made up of 27 input neurons, 7 hidden neurons and one output neuron. The 6 sets of experiments were randomly selected from orthogonal array for training and residuals were used to analyze the performance. To check the validity of regression model and to determine the significant parameter affecting the surface roughness, Analysis of variance (ANOVA) andF-test were made. The numerical analysis depict that the current to the electromagnet was an paramount parameter on surface roughness.Key words: MRAFF, ANN, Regression analysis

COMPARISON OF SUPPORT VECTOR MACHINES AND BAYESIAN NEURAL NETWORKS PERFORMANCE FOR BREAST TISSUES USING GEOSTATISTICAL FUNCTIONS IN MAMMOGRAPHIC IMAGES

2010 ◽  
Vol 09 (04) ◽  
pp. 271-288
Author(s):  
GERALDO BRAZ JUNIOR ◽  
LEONARDO DE OLIVEIRA MARTINS ◽  
ARISTÓFANES CORREA SILVA ◽  
ANSELMO CARDOSO PAIVA
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Region Of Interest ◽  
Female Breast Cancer ◽  
Support Vector ◽  
Risk Minimization ◽  
Bayesian Neural Network ◽  
Cad System ◽  
Mammographic Images ◽  
Breast Tissues

Female breast cancer is a major cause of deaths in occidental countries. Computer-aided Detection (CAD) systems can aid radiologists to increase diagnostic accuracy. In this work, we present a comparison between two classifiers applied to the separation of normal and abnormal breast tissues from mammograms. The purpose of the comparison is to select the best prediction technique to be part of a CAD system. Each region of interest is classified through a Support Vector Machine (SVM) and a Bayesian Neural Network (BNN) as normal or abnormal region. SVM is a machine-learning method, based on the principle of structural risk minimization, which shows good performance when applied to data outside the training set. A Bayesian Neural Network is a classifier that joins traditional neural networks theory and Bayesian inference. We use a set of measures obtained by the application of the semivariogram, semimadogram, covariogram, and correlogram functions to the characterization of breast tissue as normal or abnormal. The results show that SVM presents best performance for the classification of breast tissues in mammographic images. The tests indicate that SVM has more generalization power than the BNN classifier. BNN has a sensibility of 76.19% and a specificity of 79.31%, while SVM presents a sensibility of 74.07% and a specificity of 98.77%. The accuracy rate for tests is 78.70% and 92.59% for BNN and SVM, respectively.

Sign in / Sign up

Export Citation Format

Share Document