The document presents a novel diffusion wavelet approach for 3D model matching. It combines diffusion maps with wavelets to extract multi-scale shape features from 3D models. Fisher's discriminant ratio is used to select discriminative wavelet coefficients for model representation. Models are retrieved by comparing their representation vectors at different wavelet scales. Results show the diffusion wavelet approach outperforms spherical harmonics and wavelets for 3D model retrieval.
발표자: 배재성(KAIST 석사과정)
발표일: 2018.10.
최근 딥러닝을 이용한 방법은 다양한 음성 인식 과제에서 괄목할 만한 성과를 내고 있습니다. 특히 Convolutional Neural Network (CNN)을 이용한 방식은 지역적인 특징 (local feature)들을 효과적으로 잡아낼 수 있기 때문에 비교적 짧은 시간 의존도를 가지는 음성 키워드 인식이나 음소 단위 인식과 같은 과제들에서 활발히 사용되고 있습니다. 그러나 CNN은 낮은 레벨의 특징들 간의 공간적 관계성을 고려하지 않는다는 한계점이 있습니다. 이를 극복하기 위해 캡슐 네트워크 구조를 도입하여 음성 스펙트로그램에서 추출된 특징들의 공간적 관계성을 고려하고자 하였습니다. 구글 음성 단어 데이터셋에서 CNN과 그 성능을 비교해 보았으며, 깨끗한 환경과 잡음 환경 모두에서 주목할만한 성능 향상을 이끌어 냈습니다.
Deep learning for image super resolutionPrudhvi Raj
Using Deep Convolutional Networks, the machine can learn end-to-end mapping between the low/high-resolution images. Unlike traditional methods, this method jointly optimizes all the layers of the image. A light-weight CNN structure is used, which is simple to implement and provides formidable trade-off from the existential methods.
AI&BigData Lab 2016. Александр Баев: Transfer learning - зачем, как и где.GeeksLab Odessa
4.6.16 AI&BigData Lab
Upcoming events: goo.gl/I2gJ4H
Поговорим об одной из базовых практических техник обучения нейронных сетей - предобучение, finetuning, transfer learning. В каких случаях применять, какие модели использовать, где их брать и как адаптировать.
HardNet: Convolutional Network for Local Image DescriptionDmytro Mishkin
We introduce a novel loss for learning local feature descriptors which is inspired by the Lowe's matching criterion for SIFT. We show that the proposed loss that maximizes the distance between the closest positive and closest negative patch in the batch is better than complex regularization methods; it works well for both shallow and deep convolution network architectures. Applying the novel loss to the L2Net CNN architecture results in a compact descriptor -- it has the same dimensionality as SIFT (128) that shows state-of-art performance in wide baseline stereo, patch verification and instance retrieval benchmarks. It is fast, computing a descriptor takes about 1 millisecond on a low-end GPU.
Learning Convolutional Neural Networks for GraphsMathias Niepert
This document discusses a method called Patchy for applying convolutional neural networks to graph-structured data. Patchy selects node sequences from graphs using centrality measures and assembles neighborhoods around the nodes. The neighborhoods are normalized and used as receptive fields for a convolutional architecture. Experiments on graph classification benchmarks show Patchy can outperform graph kernels in terms of efficiency and effectiveness while also supporting visualization of learned edge filters. Potential limitations include increased risk of overfitting on small datasets compared to graph kernels.
Clustering algorithms are used to group similar data points together. K-means clustering aims to partition data into k clusters by minimizing distances between data points and cluster centers. Hierarchical clustering builds nested clusters by merging or splitting clusters based on distance metrics. Density-based clustering identifies clusters as areas of high density separated by areas of low density, like DBScan which uses parameters of minimum points and epsilon distance.
A comprehensive tutorial on Convolutional Neural Networks (CNN) which talks about the motivation behind CNNs and Deep Learning in general, followed by a description of the various components involved in a typical CNN layer. It explains the theory involved with the different variants used in practice and also, gives a big picture of the whole network by putting everything together.
Next, there's a discussion of the various state-of-the-art frameworks being used to implement CNNs to tackle real-world classification and regression problems.
Finally, the implementation of the CNNs is demonstrated by implementing the paper 'Age ang Gender Classification Using Convolutional Neural Networks' by Hassner (2015).
발표자: 배재성(KAIST 석사과정)
발표일: 2018.10.
최근 딥러닝을 이용한 방법은 다양한 음성 인식 과제에서 괄목할 만한 성과를 내고 있습니다. 특히 Convolutional Neural Network (CNN)을 이용한 방식은 지역적인 특징 (local feature)들을 효과적으로 잡아낼 수 있기 때문에 비교적 짧은 시간 의존도를 가지는 음성 키워드 인식이나 음소 단위 인식과 같은 과제들에서 활발히 사용되고 있습니다. 그러나 CNN은 낮은 레벨의 특징들 간의 공간적 관계성을 고려하지 않는다는 한계점이 있습니다. 이를 극복하기 위해 캡슐 네트워크 구조를 도입하여 음성 스펙트로그램에서 추출된 특징들의 공간적 관계성을 고려하고자 하였습니다. 구글 음성 단어 데이터셋에서 CNN과 그 성능을 비교해 보았으며, 깨끗한 환경과 잡음 환경 모두에서 주목할만한 성능 향상을 이끌어 냈습니다.
Deep learning for image super resolutionPrudhvi Raj
Using Deep Convolutional Networks, the machine can learn end-to-end mapping between the low/high-resolution images. Unlike traditional methods, this method jointly optimizes all the layers of the image. A light-weight CNN structure is used, which is simple to implement and provides formidable trade-off from the existential methods.
AI&BigData Lab 2016. Александр Баев: Transfer learning - зачем, как и где.GeeksLab Odessa
4.6.16 AI&BigData Lab
Upcoming events: goo.gl/I2gJ4H
Поговорим об одной из базовых практических техник обучения нейронных сетей - предобучение, finetuning, transfer learning. В каких случаях применять, какие модели использовать, где их брать и как адаптировать.
HardNet: Convolutional Network for Local Image DescriptionDmytro Mishkin
We introduce a novel loss for learning local feature descriptors which is inspired by the Lowe's matching criterion for SIFT. We show that the proposed loss that maximizes the distance between the closest positive and closest negative patch in the batch is better than complex regularization methods; it works well for both shallow and deep convolution network architectures. Applying the novel loss to the L2Net CNN architecture results in a compact descriptor -- it has the same dimensionality as SIFT (128) that shows state-of-art performance in wide baseline stereo, patch verification and instance retrieval benchmarks. It is fast, computing a descriptor takes about 1 millisecond on a low-end GPU.
Learning Convolutional Neural Networks for GraphsMathias Niepert
This document discusses a method called Patchy for applying convolutional neural networks to graph-structured data. Patchy selects node sequences from graphs using centrality measures and assembles neighborhoods around the nodes. The neighborhoods are normalized and used as receptive fields for a convolutional architecture. Experiments on graph classification benchmarks show Patchy can outperform graph kernels in terms of efficiency and effectiveness while also supporting visualization of learned edge filters. Potential limitations include increased risk of overfitting on small datasets compared to graph kernels.
Clustering algorithms are used to group similar data points together. K-means clustering aims to partition data into k clusters by minimizing distances between data points and cluster centers. Hierarchical clustering builds nested clusters by merging or splitting clusters based on distance metrics. Density-based clustering identifies clusters as areas of high density separated by areas of low density, like DBScan which uses parameters of minimum points and epsilon distance.
A comprehensive tutorial on Convolutional Neural Networks (CNN) which talks about the motivation behind CNNs and Deep Learning in general, followed by a description of the various components involved in a typical CNN layer. It explains the theory involved with the different variants used in practice and also, gives a big picture of the whole network by putting everything together.
Next, there's a discussion of the various state-of-the-art frameworks being used to implement CNNs to tackle real-world classification and regression problems.
Finally, the implementation of the CNNs is demonstrated by implementing the paper 'Age ang Gender Classification Using Convolutional Neural Networks' by Hassner (2015).
The document provides an overview of convolutional neural networks (CNNs) presented by Junho Cho. It discusses the basic components of CNNs including convolution, pooling, rectified linear units (ReLU), and fully connected layers. It also reviews popular CNN architectures such as LeNet, AlexNet, VGGNet, GoogLeNet, and ResNet. The document emphasizes that CNNs are powerful due to their ability to learn local invariance through the use of convolutional filters and sharing weights, while also having fewer parameters than fully connected networks to prevent overfitting. Finally, it provides code examples for implementing CNN models in TensorFlow.
This document summarizes a technical seminar on using convolutional neural networks for P300 detection in brain-computer interfaces. The seminar covers an introduction to brain-computer interfaces and the P300 signal, describes existing P300 detection systems and the convolutional neural network approach, and presents the network architecture, learning process, evaluation results on two datasets showing improved detection rates over other methods, and conclusions. The seminar demonstrates that the convolutional neural network approach outperforms existing methods for P300 detection, especially with a limited number of electrodes or training epochs.
This document discusses and compares different thresholding techniques for image denoising using wavelet transforms. It introduces the concept of image denoising using wavelet transforms, which involves applying a forward wavelet transform, estimating clean coefficients using thresholding, and applying the inverse transform. It then describes several common thresholding methods - hard, soft, universal, improved, Bayes shrink, and neigh shrink. Simulation results on test images corrupted with additive white Gaussian noise show that the proposed improved thresholding technique achieves lower MSE and higher PSNR than the universal hard thresholding method, demonstrating better noise removal performance while preserving image details.
The document summarizes Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). It discusses how CNNs use kernels and pooling to extract features from images while reducing parameters. It provides examples of CNN architectures and visualizations of weights and activations. RNNs are described as allowing input/output sequences, with LSTMs addressing the vanishing gradient problem. Applications discussed include image captioning using CNN features with an RNN generator.
The mean shift procedure is a general nonparametric technique for analyzing complex multimodal feature spaces and delineating arbitrarily shaped clusters. It works by recursively finding the nearest stationary point of the underlying density function, which corresponds to the mode of the density. The mean shift procedure relates to kernel density estimation and robust M-estimators of location. It provides a versatile tool for feature space analysis that can solve many low-level computer vision tasks with few parameters.
convolutional neural network (CNN, or ConvNet)RakeshSaran5
This presentation provides an overview of Convolutional Neural Networks (CNNs). It begins with an introduction to CNNs and their advantages over fully connected networks for image recognition. It then describes the key components of a CNN, including convolution layers, ReLU layers, pooling layers, and fully connected layers. Examples of each component are provided. The presentation concludes with a discussion of CNN use cases for image recognition.
Scene classification using Convolutional Neural Networks - Jayani WithanawasamWithTheBest
The document discusses scene classification using convolutional neural networks (CNNs). It begins with an outline of the topic, then provides background on computer vision as an AI problem and the importance and challenges of scene classification. It introduces CNNs as a deep learning technique for visual pattern recognition, describing their hierarchical organization and components like convolution and pooling layers. The document also discusses traditional machine learning approaches versus deep learning for scene classification and frameworks like Caffe that can be used to implement CNNs.
Modern Convolutional Neural Network techniques for image segmentationGioele Ciaparrone
Recently, Convolutional Neural Networks have been successfully applied to image segmentation tasks. Here we present some of the most recent techniques that increased the accuracy in such tasks. First we describe the Inception architecture and its evolution, which allowed to increase width and depth of the network without increasing the computational burden. We then show how to adapt classification networks into fully convolutional networks, able to perform pixel-wise classification for segmentation tasks. We finally introduce the hypercolumn technique to further improve state-of-the-art on various fine-grained localization tasks.
The document discusses neural networks and how they can be viewed as functions. It describes how neural networks take input data and produce output predictions or classifications. The document outlines how neural networks have a layered structure where each layer is a function, and how the layers are composed together. It explains that neurons are the basic units of computation in each layer and how they operate. The document also discusses how neural network training works by optimizing the weights and biases in each layer to minimize error, and how matrix operations in neural networks can benefit from parallel processing on GPUs.
Self-Organizing Maps (SOM) are a type of neural network that can be used for clustering and visualizing complex, high-dimensional data. SOM reduces dimensionality while preserving topological relationships. It arranges nodes on a grid such that similar input vectors are mapped to nearby nodes. During training, the best matching node and its neighbors are adjusted to better match the input. This results in a 2D map where similar data clusters together. For example, a SOM was used to cluster countries based on quality of life indicators, grouping those with similar living standards. SOM can be useful for applications like data mining, pattern recognition, and more.
This document summarizes research using neuroevolution techniques like HyperNEAT to train deep learning networks on image classification tasks. It describes using HyperNEAT both to directly train networks to classify MNIST handwritten digits, and to act as a feature extractor by evolving the first layers of a network and then training subsequent layers with backpropagation. The experiments compare different HyperNEAT architectures - traditional ANNs versus convolutional networks - and evaluate their performance on classifying MNIST test images both with and without the additional backpropagation training of later layers.
Convolutional neural networks (CNNs) are made up of layers that transform input volumes to output volumes. Key layers include convolutional layers that apply filters to input volumes, pooling layers that reduce spatial size, and fully-connected layers. CNNs have been very successful for computer vision tasks due to properties like shared weights, which reduce the number of parameters compared to fully-connected networks. Example networks include LeNet, AlexNet, and ResNet, with newer models featuring deeper architectures and techniques like inception modules and residual connections.
This document provides an overview of convolutional neural networks (CNNs). It describes that CNNs are a type of deep learning model used in computer vision tasks. The key components of a CNN include convolutional layers that extract features, pooling layers that reduce spatial size, and fully-connected layers at the end for classification. Convolutional layers apply learnable filters in a local receptive field, while pooling layers perform downsampling. The document outlines common CNN architectures, such as types of layers, hyperparameters like stride and padding, and provides examples to illustrate how CNNs work.
Explores the type of structure learned by Convolutional Neural Networks, the applications where they're most valuable and a number of appropriate mental models for understanding deep learning.
The document summarizes radial basis function (RBF) networks. Key points:
- RBF networks use radial basis functions as activation functions and can universally approximate continuous functions.
- They are local approximators compared to multilayer perceptrons which are global approximators.
- Learning involves determining the centers, widths, and weights. Centers can be randomly selected or via clustering. Widths are usually different for each basis function. Weights are typically learned via least squares or gradient descent methods.
This document provides an outline for a presentation on convolutional neural networks on graphs. It begins with a brief history of deep learning and discusses how convolutional neural networks leverage the compositional and hierarchical nature of data like images. It then introduces spectral graph theory and defines key concepts like graphs, graph operators, and the graph Laplacian that are necessary to extend convolutional networks to non-Euclidean graph-structured data. The outline concludes by describing different approaches to defining graph convolutional networks and their applications.
This document discusses convolutional neural networks (CNNs) for graph-structured data. CNNs are traditionally designed for Euclidean data like images but not irregular graph data. The key ideas are:
1) Define convolution on graphs using graph spectral theory by representing signals in the graph Fourier domain.
2) Coarsen graphs using a balanced cut model to extract hierarchical patterns.
3) Perform fast graph pooling using a binary tree of coarsened graphs for downsampling.
This allows generalizing CNNs to any graph data with the same computational efficiency as standard CNNs. Related works on graph CNNs are also discussed.
The document summarizes improvements made in MobileNetV3 models, including using complementary search techniques to find efficient building blocks, modifying nonlinearities like h-swish to be more efficient, and improving expensive layers through techniques like removing unnecessary projections. It also describes experiments that showed MobileNetV3 models achieving better performance versus V1/V2 models on tasks like image classification, object detection, and semantic segmentation while maintaining high efficiency for mobile applications.
This document summarizes the Kano model of customer satisfaction. It introduces the model created by Dr. Noriaki Kano that classifies customer needs into three categories: must-be attributes which cause dissatisfaction if absent but don't increase satisfaction if present, performance attributes where more is better, and attractive/surprise attributes that increase satisfaction but don't cause dissatisfaction if absent. It then provides examples of how to apply the model to analyze past, present, and future features of refrigerators to prioritize customer needs and guide design decisions.
The document provides an overview of convolutional neural networks (CNNs) presented by Junho Cho. It discusses the basic components of CNNs including convolution, pooling, rectified linear units (ReLU), and fully connected layers. It also reviews popular CNN architectures such as LeNet, AlexNet, VGGNet, GoogLeNet, and ResNet. The document emphasizes that CNNs are powerful due to their ability to learn local invariance through the use of convolutional filters and sharing weights, while also having fewer parameters than fully connected networks to prevent overfitting. Finally, it provides code examples for implementing CNN models in TensorFlow.
This document summarizes a technical seminar on using convolutional neural networks for P300 detection in brain-computer interfaces. The seminar covers an introduction to brain-computer interfaces and the P300 signal, describes existing P300 detection systems and the convolutional neural network approach, and presents the network architecture, learning process, evaluation results on two datasets showing improved detection rates over other methods, and conclusions. The seminar demonstrates that the convolutional neural network approach outperforms existing methods for P300 detection, especially with a limited number of electrodes or training epochs.
This document discusses and compares different thresholding techniques for image denoising using wavelet transforms. It introduces the concept of image denoising using wavelet transforms, which involves applying a forward wavelet transform, estimating clean coefficients using thresholding, and applying the inverse transform. It then describes several common thresholding methods - hard, soft, universal, improved, Bayes shrink, and neigh shrink. Simulation results on test images corrupted with additive white Gaussian noise show that the proposed improved thresholding technique achieves lower MSE and higher PSNR than the universal hard thresholding method, demonstrating better noise removal performance while preserving image details.
The document summarizes Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). It discusses how CNNs use kernels and pooling to extract features from images while reducing parameters. It provides examples of CNN architectures and visualizations of weights and activations. RNNs are described as allowing input/output sequences, with LSTMs addressing the vanishing gradient problem. Applications discussed include image captioning using CNN features with an RNN generator.
The mean shift procedure is a general nonparametric technique for analyzing complex multimodal feature spaces and delineating arbitrarily shaped clusters. It works by recursively finding the nearest stationary point of the underlying density function, which corresponds to the mode of the density. The mean shift procedure relates to kernel density estimation and robust M-estimators of location. It provides a versatile tool for feature space analysis that can solve many low-level computer vision tasks with few parameters.
convolutional neural network (CNN, or ConvNet)RakeshSaran5
This presentation provides an overview of Convolutional Neural Networks (CNNs). It begins with an introduction to CNNs and their advantages over fully connected networks for image recognition. It then describes the key components of a CNN, including convolution layers, ReLU layers, pooling layers, and fully connected layers. Examples of each component are provided. The presentation concludes with a discussion of CNN use cases for image recognition.
Scene classification using Convolutional Neural Networks - Jayani WithanawasamWithTheBest
The document discusses scene classification using convolutional neural networks (CNNs). It begins with an outline of the topic, then provides background on computer vision as an AI problem and the importance and challenges of scene classification. It introduces CNNs as a deep learning technique for visual pattern recognition, describing their hierarchical organization and components like convolution and pooling layers. The document also discusses traditional machine learning approaches versus deep learning for scene classification and frameworks like Caffe that can be used to implement CNNs.
Modern Convolutional Neural Network techniques for image segmentationGioele Ciaparrone
Recently, Convolutional Neural Networks have been successfully applied to image segmentation tasks. Here we present some of the most recent techniques that increased the accuracy in such tasks. First we describe the Inception architecture and its evolution, which allowed to increase width and depth of the network without increasing the computational burden. We then show how to adapt classification networks into fully convolutional networks, able to perform pixel-wise classification for segmentation tasks. We finally introduce the hypercolumn technique to further improve state-of-the-art on various fine-grained localization tasks.
The document discusses neural networks and how they can be viewed as functions. It describes how neural networks take input data and produce output predictions or classifications. The document outlines how neural networks have a layered structure where each layer is a function, and how the layers are composed together. It explains that neurons are the basic units of computation in each layer and how they operate. The document also discusses how neural network training works by optimizing the weights and biases in each layer to minimize error, and how matrix operations in neural networks can benefit from parallel processing on GPUs.
Self-Organizing Maps (SOM) are a type of neural network that can be used for clustering and visualizing complex, high-dimensional data. SOM reduces dimensionality while preserving topological relationships. It arranges nodes on a grid such that similar input vectors are mapped to nearby nodes. During training, the best matching node and its neighbors are adjusted to better match the input. This results in a 2D map where similar data clusters together. For example, a SOM was used to cluster countries based on quality of life indicators, grouping those with similar living standards. SOM can be useful for applications like data mining, pattern recognition, and more.
This document summarizes research using neuroevolution techniques like HyperNEAT to train deep learning networks on image classification tasks. It describes using HyperNEAT both to directly train networks to classify MNIST handwritten digits, and to act as a feature extractor by evolving the first layers of a network and then training subsequent layers with backpropagation. The experiments compare different HyperNEAT architectures - traditional ANNs versus convolutional networks - and evaluate their performance on classifying MNIST test images both with and without the additional backpropagation training of later layers.
Convolutional neural networks (CNNs) are made up of layers that transform input volumes to output volumes. Key layers include convolutional layers that apply filters to input volumes, pooling layers that reduce spatial size, and fully-connected layers. CNNs have been very successful for computer vision tasks due to properties like shared weights, which reduce the number of parameters compared to fully-connected networks. Example networks include LeNet, AlexNet, and ResNet, with newer models featuring deeper architectures and techniques like inception modules and residual connections.
This document provides an overview of convolutional neural networks (CNNs). It describes that CNNs are a type of deep learning model used in computer vision tasks. The key components of a CNN include convolutional layers that extract features, pooling layers that reduce spatial size, and fully-connected layers at the end for classification. Convolutional layers apply learnable filters in a local receptive field, while pooling layers perform downsampling. The document outlines common CNN architectures, such as types of layers, hyperparameters like stride and padding, and provides examples to illustrate how CNNs work.
Explores the type of structure learned by Convolutional Neural Networks, the applications where they're most valuable and a number of appropriate mental models for understanding deep learning.
The document summarizes radial basis function (RBF) networks. Key points:
- RBF networks use radial basis functions as activation functions and can universally approximate continuous functions.
- They are local approximators compared to multilayer perceptrons which are global approximators.
- Learning involves determining the centers, widths, and weights. Centers can be randomly selected or via clustering. Widths are usually different for each basis function. Weights are typically learned via least squares or gradient descent methods.
This document provides an outline for a presentation on convolutional neural networks on graphs. It begins with a brief history of deep learning and discusses how convolutional neural networks leverage the compositional and hierarchical nature of data like images. It then introduces spectral graph theory and defines key concepts like graphs, graph operators, and the graph Laplacian that are necessary to extend convolutional networks to non-Euclidean graph-structured data. The outline concludes by describing different approaches to defining graph convolutional networks and their applications.
This document discusses convolutional neural networks (CNNs) for graph-structured data. CNNs are traditionally designed for Euclidean data like images but not irregular graph data. The key ideas are:
1) Define convolution on graphs using graph spectral theory by representing signals in the graph Fourier domain.
2) Coarsen graphs using a balanced cut model to extract hierarchical patterns.
3) Perform fast graph pooling using a binary tree of coarsened graphs for downsampling.
This allows generalizing CNNs to any graph data with the same computational efficiency as standard CNNs. Related works on graph CNNs are also discussed.
The document summarizes improvements made in MobileNetV3 models, including using complementary search techniques to find efficient building blocks, modifying nonlinearities like h-swish to be more efficient, and improving expensive layers through techniques like removing unnecessary projections. It also describes experiments that showed MobileNetV3 models achieving better performance versus V1/V2 models on tasks like image classification, object detection, and semantic segmentation while maintaining high efficiency for mobile applications.
This document summarizes the Kano model of customer satisfaction. It introduces the model created by Dr. Noriaki Kano that classifies customer needs into three categories: must-be attributes which cause dissatisfaction if absent but don't increase satisfaction if present, performance attributes where more is better, and attractive/surprise attributes that increase satisfaction but don't cause dissatisfaction if absent. It then provides examples of how to apply the model to analyze past, present, and future features of refrigerators to prioritize customer needs and guide design decisions.
The Kano model is a theory developed in the 1980s to help companies analyze customer needs and determine what features delight or satisfy basic needs. It categorizes needs into three types: dissatisfiers (basic needs that cause dissatisfaction if absent), satisfiers (needs where more is better), and delighters (unexpected needs that impress customers). For example, cleanliness is a basic hotel need that causes dissatisfaction if absent, while internet access satisfies as quality increases, and complimentary cookies delight customers. The model helps companies understand changing customer expectations over time to meet minimum needs and differentiate through delighters.
The document provides an overview of the Kano Model, which was developed by Noriaki Kano to challenge traditional customer satisfaction models. The Kano Model classifies customer requirements into three categories - dissatisfiers, satisfiers, and delighters - based on their impact on customer satisfaction. The document outlines the origins, key elements, methodology, and applications of the Kano Model for identifying voice of the customer needs and determining appropriate strategies.
Fuzzy logic is a flexible machine learning technique that mimics human thought by allowing intermediate values between true and false. It provides a mechanism for interpreting and executing commands based on approximate or uncertain reasoning. Unlike binary logic which can only have true or false values, fuzzy logic uses linguistic variables and degrees of membership to represent concepts that may have a partial truth. Fuzzy systems find applications in automatic control, prediction, diagnosis and user interfaces.
- Fuzzy logic was developed by Lotfi Zadeh to address applications involving subjective or vague data like "attractive person" that cannot be easily analyzed using binary logic. It allows for partial truth values between completely true and completely false.
- Fuzzy logic controllers mimic human decision making and involve fuzzifying inputs, applying fuzzy rules, and defuzzifying outputs. This allows systems to be specified in human terms and automated.
- Fuzzy logic has many applications from industrial process control to consumer products like washing machines and microwaves. It offers an intuitive way to model real-world ambiguities compared to mathematical or logic-based approaches.
Awarded presentation of my research activity, PhD Day 2011, February 23th 2011, Cagliari, Italy.
This presentation has been awarded as the best one of the track on information engineering.
Want to know more?
see my publications at
http://prag.diee.unica.it/pra/ita/people/satta
Ijri ece-01-02 image enhancement aided denoising using dual tree complex wave...Ijripublishers Ijri
This paper presents a novel way to reduce noise introduced or exacerbated by image enhancement methods, in particular algorithms based on the random spray sampling technique, but not only. According to the nature of sprays, output images of spray-based methods tend to exhibit noise with unknown statistical distribution. To avoid inappropriate assumptions on the statistical characteristics of noise, a different one is made. In fact, the non-enhanced image is considered to be either free of noise or affected by non-perceivable levels of noise. Taking advantage of the higher sensitivity of the human visual system to changes in brightness, the analysis can be limited to the luma channel of both the non-enhanced and enhanced image. Also, given the importance of directional content in human vision, the analysis is performed through the dual-tree complex wavelet transform , lanczos interpolator and edge preserving smoothing filters. Unlike the discrete wavelet transform, the DTWCT allows for distinction of data directionality in the transform space. For each level of the transform, the standard deviation of the non-enhanced image coefficients is computed across the six orientations of the DTWCT, then it is normalized.
Keywords: dual-tree complex wavelet transform (DTWCT), lanczos interpolator, edge preserving smoothing filters.
Boosting CED Using Robust Orientation Estimationijma
n this paper, Coherence Enhancement Diffusion (CED) is boosted feeding external orientation using new
robust orientation estimation. In CED, proper scale selection is very important as the gradient vector at
that scale reflects the orientation of local ridge. For this purpose a new scheme is proposed in which pre
calculated orientation, by using local and integration scales. From the experiments it is found the proposed
scheme is working much better in noisy environment as compared to the traditional Coherence
Enhancement Diffusion
Ijri ece-01-02 image enhancement aided denoising using dual tree complex wave...Ijripublishers Ijri
This paper presents a novel way to reduce noise introduced or exacerbated by image enhancement methods, in particular
algorithms based on the random spray sampling technique, but not only. According to the nature of sprays,
output images of spray-based methods tend to exhibit noise with unknown statistical distribution. To avoid inappropriate
assumptions on the statistical characteristics of noise, a different one is made. In fact, the non-enhanced image is
considered to be either free of noise or affected by non-perceivable levels of noise. Taking advantage of the higher sensitivity
of the human visual system to changes in brightness, the analysis can be limited to the luma channel of both the
non-enhanced and enhanced image. Also, given the importance of directional content in human vision, the analysis is
performed through the dual-tree complex wavelet transform , lanczos interpolator and edge preserving smoothing filters.
Unlike the discrete wavelet transform, the DTWCT allows for distinction of data directionality in the transform space.
For each level of the transform, the standard deviation of the non-enhanced image coefficients is computed across the
six orientations of the DTWCT, then it is normalized.
Keywords: dual-tree complex wavelet transform (DTWCT), lanczos interpolator, edge preserving smoothing filters.
Boosting ced using robust orientation estimationijma
In this paper, Coherence Enhancement Diffusion (CED) is boosted feeding external orientation using new
robust orientation estimation. In CED, proper scale selection is very important as the gradient vector at
that scale reflects the orientation of local ridge. For this purpose a new scheme is proposed in which pre
calculated orientation, by using local and integration scales. From the experiments it is found the proposed
scheme is working much better in noisy environment as compared to the traditional Coherence
Enhancement Diffusion
A Simple Introduction to Neural Information RetrievalBhaskar Mitra
Neural Information Retrieval (or neural IR) is the application of shallow or deep neural networks to IR tasks. In this lecture, we will cover some of the fundamentals of neural representation learning for text retrieval. We will also discuss some of the recent advances in the applications of deep neural architectures to retrieval tasks.
(These slides were presented at a lecture as part of the Information Retrieval and Data Mining course taught at UCL.)
1. The document presents a new approach for steganography detection using a combination of Fisher's linear discriminant function (FLD) and radial basis function neural network (RBF).
2. In the training phase, FLD is used to project high-dimensional image data onto a lower dimensional space, then an RBF network is trained to classify images as containing hidden data or not.
3. Experiments show the combined FLD-RBF method provides promising results for steganography detection compared to existing supervised methods, though extracting the hidden information remains challenging.
Improvement of Anomaly Detection Algorithms in Hyperspectral Images Using Dis...sipij
Recently anomaly detection (AD) has become an important application for target detection in hyperspectral remotely sensed images. In many applications, in addition to high accuracy of detection we need a fast and reliable algorithm as well. This paper presents a novel method to improve the performance of current AD algorithms. The proposed method first calculates Discrete Wavelet Transform (DWT) of every pixel vector of image using Daubechies4 wavelet. Then, AD algorithm performs on four bands of “Wavelet transform” matrix which are the approximation of main image. In this research some benchmark AD algorithms including Local RX, DWRX and DWEST have been implemented on Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) hyperspectral datasets. Experimental results demonstrate significant improvement of runtime in proposed method. In addition, this method improves the accuracy of AD algorithms because of DWT’s power in extracting approximation coefficients of signal, which contain the main behaviour of signal, and abandon the redundant information in hyperspectral image data.
Getting started with chemometric classificationAlex Henderson
The document provides an overview of chemometric classification and resources for working with spectroscopic data. It discusses key terminology like variables, observations, and vector space. It also covers important preprocessing steps like normalization, mean centering, and principal components analysis (PCA). PCA finds orthogonal principal components that maximize the explained variance in the data in a lower dimensional space.
An efficient fuzzy classifier with feature selection basedssairayousaf
This document presents an efficient fuzzy classifier with feature selection capabilities. A fuzzy entropy measure is used to partition the input feature space into non-overlapping decision regions and to select relevant features. Fuzzy entropy evaluates the information of pattern distribution in the pattern space. The decision regions do not overlap, reducing computational complexity and load. Classification speed is extremely fast while still achieving good performance by correctly determining decision region boundaries. Feature selection via fuzzy entropy reduces dimensionality by discarding noisy, redundant, and unimportant features. The proposed classifier is applied to two databases with good classification results, demonstrating its effectiveness.
This document discusses dimensionality reduction techniques for hyperspectral images. It proposes using k-means clustering based on statistical measures like variance, standard deviation, and mean absolute deviation to select bands from hyperspectral images. The number of bands is first estimated using virtual dimensionality. Bands are then clustered based on their statistical properties and one band is selected from each cluster with the maximum value of the statistical measure. Finally, endmembers are extracted from the selected bands using N-FINDR.
Mimo radar detection in compound gaussian clutter using orthogonal discrete f...ijma
This paper proposes orthogonal Discrete Frequency Coding Space Time Waveforms (DFCSTW) for
Multiple Input and Multiple Output (MIMO) radar detection in compound Gaussian clutter. The proposed
orthogonal waveforms are designed considering the position and angle of the transmitting antenna when
viewed from origin. These orthogonally optimized show good resolution in spikier clutter with Generalized
Likelihood Ratio Test (GLRT) detector. The simulation results show that this waveform provides better
detection performance in spikier Clutter.
This document discusses challenges in analyzing large spatiotemporal environmental datasets and presents hierarchical Gaussian process models as a framework to address these challenges. It describes a joint NASA and Forest Service initiative applying these models to map forest variables in Interior Alaska using sparse ground plots and airborne/spaceborne LiDAR data. The key challenges are incorporating diverse spatial data sources, handling missingness/misalignment, accounting for spatial dependence, propagating uncertainty, and scaling to massive datasets.
This document summarizes a research paper that designed and implemented sphere decoding (SD) for multiple-input multiple-output (MIMO) systems using an FPGA. It used Newton's iterative method to calculate the matrix inverse as part of the SD algorithm, which reduces complexity compared to direct matrix inversion. The authors implemented SD for a 2x2 MIMO system with 4-QAM modulation. Simulation results showed that Newton's method converged after 7 iterations, and SD successfully calculated the minimum Euclidean distance vector.
This document introduces the Continuous Crooklet Transform (CCrT) and Discrete Crooklet Transform (DCrT) as novel transforms that aim to overcome limitations of wavelet and curvelet transforms. The CCrT is defined as the convolution of an input signal with scaled and translated versions of a principal "crooklet" function. The DCrT decomposes a signal using a filter bank of low-pass and high-pass filters in a manner similar to curvelet transforms. The Crooklet Transform fits image properties better than wavelets and has less computational complexity than curvelets. Potential applications of the Crooklet Transform include image processing, feature extraction, image compression, and medical imaging.
This document summarizes a research paper that proposes a novel architecture for implementing a 1D lifting integer wavelet transform (IWT) using residue number system (RNS). The key aspects covered are:
1) RNS offers advantages over binary representations for digital signal processing by avoiding carry propagation. A ROM-based approach is proposed for RNS division.
2) The lifting scheme for discrete wavelet transforms is summarized, including split, predict, and update stages.
3) A novel RNS-based architecture is proposed using three main blocks - split, predict, and update - that repeat at each decomposition level. Pipelined implementations of the predict and update blocks are detailed.
The document summarizes research on mesh representations in computer graphics. It discusses Greg Turk's introduction of "mutual tessellation" to represent objects at different levels of detail. It also covers Hugues Hoppe's work on "mesh optimization" to minimize triangles in dense meshes and "progressive meshes" to preserve overall appearance while simplifying. The document outlines challenges of mesh simplification, level-of-detail approximations, and compression. It describes representing meshes as sets of vertices, connectivity, and attributes. An energy function is defined to optimize meshes by minimizing distance and spring energies while preserving scalar attributes. Applications in medical imaging and reduced manual work in graphics are mentioned.
[DSC Adria 23]Davor Horvatic Human-Centric Explainable AI In Time Series Anal...DataScienceConferenc1
To fully trust, accept, and adopt newly emerging AI solutions in our everyday lives and practices, we need human-centric explainable AI that can provide human-understandable interpretations for their algorithmic behaviour and outcomes—consequently enabling us to control and continuously improve their performance, robustness, fairness, accountability, transparency, and explainability throughout the entire lifecycle of AI applications. The recently emerging trend within diverse and multidisciplinary research forms the basis of the next wave of AI. In this talk, we will present research that plans to produce interpretable deep learning models for time series analysis with a broad scope of applications.
This document summarizes a research paper that proposes using multi-objective particle swarm optimization (MOPSO) to maximize coverage and lifetime in wireless sensor networks. The paper formulates the sensor network optimization problem as maximizing two objectives: coverage of the monitored area and lifetime of the sensor network. It describes using MOPSO to determine optimal sensor node locations that provide a set of Pareto optimal trade-offs between these two objectives. Simulation results show MOPSO obtaining a well-populated Pareto front of solutions and example optimal layouts that maximize either coverage or lifetime.
Similar to A Diffusion Wavelet Approach For 3 D Model Matching (20)
A Diffusion Wavelet Approach For 3 D Model Matching
1. A Diffusion Wavelet Approach for 3-D Model Matching Authors: K.P. Zhu, Y.S. Wong, W.F. Lu, J.Y.H. Fuh Presented by: Raphael Steinberg
2. Schedule Introduction Diffusion Maps Wavelets and Diffusion Wavelets Fisher’s Discriminant Ratio (FDR) Retrieval Procedure Results Conclusions 2
3. Introduction Currently - A larger than ever number of 3D Models in CAD, computer games, multimedia, molecular biology, computer vision and more There is a need for 3D Retrieval 3
4. Introduction (2) Tagging are not always available or sufficient to describe the model we require Combine topological information with multi-scale properties 4
5. Model Reusability (CAD/Animation) Model Matching Video Retrieval (2.5D/Virtual environments) Ecommerce Correcting defects Efficient Representation Many other uses… Motivation for 3D Retrieval 5
6. Obstacles in Retrieval Partial retrieval - Non-transitive Functional description How to match text tags with vertices and texture? Orthonormal coordinate system 6
8. 3D Model Matching – Prior Art Feature vectors using wavelets to mesh vertices – localized in both space & frequency – Paquet et. al. 2000 Random sampling for comparison – Osada et. al. 2001 8
9. Spherical harmonics (SH) Global method in Euclidean space lacks multi-scale analysis Legendre polynomials solve the Laplace equation in Spherical coordinates Vranic et. al. 2001 9
10. Spherical Wavelets (SW) Multi-scale in Euclidean space Lacks connectivity on the manifold Tannenbaum et. al. 2007 10
11. Schedule Introduction Diffusion Maps Wavelets and Diffusion Wavelets Fisher’s Discriminant Ratio (FDR) Retrieval Procedure Results Conclusions 11
12. Diffusion Maps Introduction Originally suggested by Stephan Lafon and R.R. Coifman from Yale Math, circa 2005 Many other manifold learning techniques exist Data analysis based on geometric properties of the data set 12
20. Use RBF Gaussian Kernel to choose ε Normalize W to create a Stochastic Matrix Diffusion Maps Algorithm 16 Lu et. al. 2009
21. Diffusion Maps algorithm (2) Diffuse by taking higher powers of t “The diffusion distance is equal to the Euclidean distance in the diffusion map space” , Nadler et. al. 2005 Cut manifold according to dominant eigenvalues 17
22. Diffusion Maps Code Example function checker(); close all; tetha=2*pi*rand(1,500); z=[cos(tetha);sin(tetha)]; figure(1);scatter(z(1,:),z(2,:),'b*');hold on; N=size(z,2); epsilon=linspace(0.01,.3,10); %epsilon=.3; W=nan(N); summer=nan(1,length(epsilon)); for k=1:length(epsilon) for i=1:N parfor j=1:N W(i,j)=exp(-sum((z(:,j)-z(:,i)).^2)/2/epsilon(k)); end end summer(k)=sum(sum(W)); end figure;scatter(log(epsilon),log(summer));title('Epsilon - linear region') p=polyfit(log(epsilon),log(summer),1); d=2*p(1);%manifold dimension M=W*diag(1./sum(W,2)); [U V]=svds(M); sync=max(U(:,2)); figure(1);scatter(U(:,2)./sync,U(:,3)./sync,'rd') title('Original manifold as stars and reconstructed manifold as diamonds') end 18
23. Schedule Introduction Diffusion Maps Wavelets and Diffusion Wavelets Fisher’s Discriminant Ratio (FDR) Retrieval Procedure Results Conclusions 19
26. Novelty – Diffusion Wavelets Combination of Diffusion Maps and Wavelets Used for non-linear dimensionality reduction Extension of wavelets to the unit circle (just as diffusion maps extends the Fourier transform) 22
28. Example of Diffusion Wavelets 24 Wavelet basis ψ(2,2,3) Scaling basis φ(1,1,1) Wavelet basis ψ(4,2,5) Wavelet basis ψ(3,2,3)
29. Diffusion Wavelets Use an optimization scheme to construct the scaling functions Each scaling function should deal with a single dimension and be orthogonal to the other scaling functions Extension of wavelets to the sphere (or to any other manifold) 25
30. Diffusion Wavelets (2) Better than LOD (Level of Detail - simplifies meshes) Involved algorithm – very few implementations exist 26
37. IRPR Curve Measure performance – use Princeton University 3D database IRPR – Information Retrieval Precision-Recall 33
38. IRPR Curve m = relevant matches r = # of retrieved models 1) Precision = 2) Recall = 34
39. Schedule Introduction Diffusion Maps Wavelets and Diffusion Wavelets Fisher’s Discriminant Ratio (FDR) Retrieval Procedure Results Conclusions 35
40. 3D Model Retrieval Procedure Compute the diffusionwavelet for each 3D model Obtain the model representing vector X Compute the 2nd order statistics of X for each scale 36
41. 1) Start with a coarsest scale comparison 2)Advance up to the finest scale 3) Stop on threshold or when finest scale reached * Use a threshold to determine if a model is from a certain class Model Matching Procedure 37
42. Schedule Introduction Diffusion Maps Wavelets and Diffusion Wavelets Fisher’s Discriminant Ratio (FDR) Retrieval Procedure Results Conclusions 38
43. Experimental Results 39 Differences in scaling levels DW gives better results than SH and SW
45. Schedule Introduction Diffusion Maps Wavelets and Diffusion Wavelets Fisher’s Discriminant Ratio (FDR) Retrieval Procedure Results Conclusions 41
46. Authors’ Conclusions Surfaces with sharp peaks, grooves or holes contain high-frequency information which is not addressed by the wavelet multi-resolution (use diffusion wavelet packets instead?) Possible to extend to partial matching DW presents better results than SH and SW 42
52. Seems like a reasonable solution to the problem of 3D object retrieval44
53.
54. How are the wavelet functions affected when a new model is inserted?45
55.
56.
57. Can we have an extension of Diffusion Wavelets for non-rigid manifolds?47
58. “Would like to have” (Technical/2) How to automatically choose the level of decomposition 48
59. “Would like to have” (Technical/3) An intuitive explanation - why prefer Diffusion Wavelets over Diffusion Wavelet Packets? Wavelet Packets seem to give more information especially in high frequencies… 49
60. “Would like to have” (Technical/4) Numerical problems of overflow of the FDR - use logarithm instead of inverse? 50
61. “Would like to have” (Presentation/1) Block diagram of the algorithm 51
62. “Would like to have” (Presentation/2) Web-based Graphical User Interface 52
64. “Would like to have” (Presentation/4) More explanations on Diffusion Wavelets 54
65. Conclusions Shape retrieval requires multi-scale analysis 3D models, like most real-life objects, are embedded in a low dimension manifold Results are robust to noise and to mesh simplifications 55
66. Conclusions (2) Diffusion Wavelets give good retrieval results for 3D objects Possible to extend the proposed method to include texture, sound, smell, elasticity and any other possibly given attribute of the 3D model 56
68. References [1]K.P. Zhu, Y.S. Wong, W.F. Lu, J.Y.H. Fuh. , Department of Mechanical Engineering, National University of Singapore “A diffusion wavelet approach for 3-D model matching” Computer Aided Design, Elsevier, Nov. 2008 [2] Presentation by R.R. Coifman et. al. [3] J. Lu et. al. “Dominant Texture and Diffusion Distance Manifolds“, Eurographics, Volume 28 , Issue 2, Pages 667 - 676, Mar. 2009 [4] Diffusion waveletsMatlab code: http://www.math.duke.edu/~mauro/diffusionwavelets.html#Code|outline [5] The Princeton Shape Benchmark: http://shape.cs.princeton.edu/benchmark/ [6] Nadler, B., Lafon, S., Coifman, R., Kevrekidis, I. “Diffusion maps, spectral clustering and eigenfunctions of Fokker-Planck operators”. [7] Ulrike von Luxburg, “A tutorial on spectral clustering”. Statistical Journal 2007 [8] Personal communications with K.P. Zhu [9] MANI - Manifold learning Matlab tool http://www.math.umn.edu/~wittman/mani/ [10] Vranic D, Saupe D, Richter J. Tools for 3D-object retrieval: Karhunen-Loeve transform and spherical harmonics. In: Proc. IEEE workshop on multimedia signal processing; 2001. p. 29398. 58
69. References [11] Osada R, Funkhouser T, Chazelle B, Dobkin D. Matching 3D models with shape distributions, In: Proc. shape modeling international. 2001. p. 15466. [12] Laga H, Nakajima M. Statistical spherical wavelet moments for content-based 3D model Retrieval. In: Computer graphics international 2007, CGI. 2007; 2007. p.1-8. [13] Nain D, Haker S, Bobick A, Tannenbaum A. Multiscale 3-D shape representation and segmentation using spherical wavelets. IEEE Transactions on Medical Imaging 2007;26(4), pages 598-618. 59
Editor's Notes
Google has their own format –SketchUp (SU) and has been investing a lot of effort in 3D technologies. More examples from google include – google earth, google sketch-up and O3D platform for browser 3D display
Other manifold learning techniques – PCA (+variants), Kernel PCA, LLE (locally linear embedding), ISOMAP, Hessian LLE, LaplacianEigenMaps and many more…
Reference[9]
Reference [2]
See [1,6] for diffusion maps equations and [3] for graph
Ignore λmax=1 since it is a trivial eigenvalue of the markov transition matrix and corresponds to the stationary distribution of the markov chain at t=∞Cut is promised to conform with min-cut max-flow algorithmSee reference [6] for a thorough description of diffusion maps
We see that it is possible to reconstruct the manifold using just a single eigenvalue. Cuts can be made on the eigenvectors that represent the manifold (in this case, the second or the third eigenvectors corresponding to the second or third largest eigenvalue) – these cuts are meaningful since they are taking into account the geometric distribution of the original points. We see that diffusion maps approximate the Fourier series over the circle as the sine and cosine functions are the solution of the differential equation f’’=-f
Animated objects can be more sensitive to mesh simplification algorithms than CAD models.
Haar wavelet with 3 levels of decomposition to the Stanford Bunny image. By applying a threshold in the wavelet domain we can efficiently find similar images. The threshold is a very efficient way to remove noise. The high value coefficients correspond to edges at various scales. Collecting the high value coefficients to create a feature vector would ensure a good representation of the image. For example, check the lossy compression algorithm JPEG-2000.
We take 4 levels of decomposition since there is no real advantage in taking more decomposition levels and the computational burden is heavy. This result is specific to Princeton’s database and can change when dealing with different databases.
Take inverse of FDR to avoid numeric problems of overflow. Select model with minimum within cluster scattering and maximum within cluster scattering
Can do training on the entire database…See [5] for database
Chair – sparse structurePlane – smooth surface with local singularityKangaroo – smooth surfaceFlower – combines smooth surface with local singularity