检索结果-内蒙古大学图书馆

International Joint Conference on Artificial Intelligence, JCAI

作者： ZhenYu Qiu Hui Wei Department of Computer Science Laboratory of Algorithm for Cognitive Model Fudan University Shanghai China

The traditional three-dimensional object recognition method based on hypothesise and test need to solve the coordinate transformation matrix from scene to model through a group of non-linear equations. Therefore, it has a very high complexity. This paper presents a man-made object recognition method based on the geometry feature of line segments characteristics, and disperses the overall coordinate transformation calculation in every local plane homography calculation, reduces the complexity of the solution. First we pre-match the feature points using geometric invariants, then assume and solve the plane homography matrix between scenes to model. After that we match the line segments on the homography plane, and by this we verify the assumption. Experiments proved that this method can rapidly and accurately identify man-made objects which contain coplanar line segment features.

关键词： Object recognition Testing Layout Image edge detection Transmission line matrix methods Shape Robotic assembly Artificial intelligence Computer science Nonlinear equations

来源：评论

学校读者我要写书评

暂无评论

Acquiring 3D Experience of View Angle Estimation Based on Monocular and Single Image

Acquiring 3D Experience of View Angle Estimation Based on Mo...

引用

Congress on Image and Signal Processing, CISP

作者： Hui Wei Zhen-Yu Qiu Department of Computer Science Laboratory of Algorithm for Cognitive Model Fudan University Shanghai China

It has been proved that acquired training is important to the development of stereopsis experience. Month-old babies already have the initial experience of invariance recognition of 3D objects. There is a slight lack of precision in the interpretation of biological vision. However, the small cost and the fast speed in calculation meet the requirements of invariance recognition, the rich visual experience in which play an important role. But what is the experience, how to acquire and how to use, these problems have never been satisfactorily resolved. In this paper we simulate the learning of visual experience in children, and solve a view angle estimated problem by using self-organizing network, which make the hidden experience clarified. Compared to the Classic camera calibration, which a large number of parameters need to be estimated, this method needs only one image and does not aim to 3D reconstruction. By avoiding the complex calibration and registration process, an amount of computation has been reduced. Visual experiences are all obtained from the most ordinary examples, and the characterization based on the geometric feature. Therefore, this method has strong expansibility and good generalization ability.

关键词： Cameras Calibration Signal processing algorithms Pediatrics Computer vision Biomedical signal processing Computer science Biological system modeling Costs Self-organizing networks

来源：评论

学校读者我要写书评

暂无评论

3D Information Acquisition from Rectangles and Circles in Single Images

3D Information Acquisition from Rectangles and Circles in Si...

引用

Computer Science and Information Engineering, CSIE, WRI World Congress on FLAGGED

作者： Ren Yuan Wei Hui Laboratory of Algorithm of Cognitive Model School of Computer Science Fudan University Shanghai China

This paper presents a monocular approach of partially recovering three-dimensional information from rectangles and circles with proper prior knowledge. Taking image-formation process as projective transform, we first work out the focal length of an image with a common perspective rectangle and further obtain normal vectors of the supporting planes of rectangles and circles and relative depths of each point on the planes with respect to camera coordinate system. The absolute depths and other information about lengths and distances are also acquirable with reasonable estimation or measurement of side lengths and radius. The experimental results suggest good feasibility and acceptable precision of our approach.

关键词： Image reconstruction Cameras Shape Computer vision Computer science Layout Optical imaging Pixel Computational geometry Algebra

来源：评论

学校读者我要写书评

暂无评论

A small-world of neuronal functional network from multi-electrode recordings during a working memory task

A small-world of neuronal functional network from multi-elec...

引用

International Joint Conference on Neural Networks

作者： Hu Lu Bao-Ming Li Hui Wei School of Computer Science Laboratory of Cognitive and Model Algorithm Fudan University Shanghai China

ISBN: (纸本)9781467314886

Graph theory is a very useful tool in the study of functional and anatomical network in the brain. It had been widely used in the functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) signals. Only very few studies analyzed the neuronal connections composed of individually recorded neurons. Particularly applying in the case of neuronal functional networks of animal behavior-dependent was rare. Scientists have found the small-world network properties in the functional network derived from fMRI and EEG signals. Whether there are existing small-world properties in the neuronal networks of multi-electrode recording? We use graph theory techniques to construct and analyze the neuronal networks. In the functional networks of a simultaneously recorded population of neurons in prefrontal cortex of the rat, in a Y-maze working memory task, we find that the neuronal connection density is highly relevant to rat behavior. We find there is a small-world effect in the neuronal functional network compared to a random graph with the same size and average connection density. We also find that small-world properties have a great relationship to correlation coefficient threshold selection. These findings indicate that neuronal functional networks of multi-electrode recordings are also small-world networks. Network connection topology and connection density are related to the working memory tasks in the rat.

关键词： Graph theory Functional network Small-world network Y-maze Multi-electrode recording

来源：评论

学校读者我要写书评

暂无评论

A model of V4 neurons based on sparse coding

Lecture Notes in Computer Science (including subseries Lectu...

引用

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2014年 8834卷 239-246页

作者： Wei, Hui Dong, Zheng Li, Qiang Department of Computer Science Laboratory of Cognitive Model and Algorithm Shanghai Key Laboratory of Data Science Fudan University Shanghai China

Area V4 lies in the middle of the ventral visual pathway in primate brains. It is an intermediate stage in the visual processing for object discrimination. V4 neurons exhibit selectivity to complex boundary conformation. In this paper, we propose a novel model of V4 neurons based on sparse coding. The model is a multi-layer neural network of which the output layer consists of laterally connected V4 units. We provide an informal proof for sparse coding with intra-layer inhibitory connections and show experimentally that this model successfully reproduces shape selectivity observed in V4 neurons. The model provides clues to the high level representation of visual stimuli in the brain © Springer International Publishing Switzerland 2014.

关键词： Neurons

来源：评论

学校读者我要写书评

暂无评论

CTFDP: an affine invariant method for matching contours

CTFDP: an affine invariant method for matching contours

引用

International Conference on Machine Learning and Cybernetics (ICMLC)

作者： Hui-Xuan Tang Hui Wei Laboratory of Algorithm for Cognitive Model Department of Computer Science and Engineering Fudan University Shanghai China

In this paper a new method for matching contours called CTFDP is presented. It is invariant to affine transformations and can provide robust and accurate estimation of point correspondence between closed curves. This has all been achieved by exploiting the dynamic programming techniques in a coarse-to-fine framework. By normalizing the shape into a standard point distribution, the new method can compare different shapes despite the shearing and scaling effect of affine transformation. Using the coarse-to-fine dynamic programming technique, the shapes are aligned to each other by iteratively seeking for correspondences and estimating relative transformations so as to prune the start points in the dynamic programming stage in turn. Experiments on artificial and real images have validated the robustness and accuracy of the presented method.

关键词： Shape Dynamic programming Robustness Noise shaping Computational efficiency Computer science Electronic mail Shearing Computer vision Application software

来源：评论

学校读者我要写书评

暂无评论

V4 Neural Network model for Visual Saliency and Discriminative Local Representation of Shapes

V4 Neural Network Model for Visual Saliency and Discriminati...

引用

International Joint Conference on Neural Networks

作者： Hui Wei Zheng Dong Department of Computer Science Laboratory of Cognitive Model and Algorithm Fudan University Shanghai Key Laboratory of Data Science Fudan University

ISBN: (纸本)9781479914821

Visual area V4 lies in the middle of the ventral visual pathway in the primate brain. It is an intermediate stage in the visual processing for object discrimination. It plays an important role in the neural mechanism of visual attention and shape recognition. V4 neurons exhibit selectivity for salient features of contour conformation. In this paper, we propose a novel model of V4 neurons based on a multilayer neural network inspired by recent studies on V4. Its low-level layers consist of computational units simulating simple cells and complex cells in the primary visual cortex. These layers extract preliminary visual features including edges and orientations. The V4 computational units calculate the entropy of the extracted features as a measure of visual saliency. The salient features are then selected and encoded with a layer of Restricted Boltzmann Machine to generate an intermediate representation of object shapes. The model was evaluated in shape distinction, handwritten digits classification, feature detection, and feature matching experiments. The results demonstrate that this model generates discriminative local representation of object shapes. It provides clues to understand the high level representation of visual stimuli in the brain.

关键词： Shape Neurons Visualization Entropy Computational modeling Brain modeling Gratings

来源：评论

学校读者我要写书评

暂无评论

A neural circuit model for nCRF's dynamic adjustment and its application on image representation

A neural circuit model for nCRF's dynamic adjustment and its...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Hui Wei Xiao-Mei Wang Department of Computer Science Laboratory of Cognitive Model and Algorithm Brain Science Research Center Fudan University Shanghai China

According to Biology there is a large disinhibitory area outside the classical receptive field (CRF), which is called as non-classical receptive field (nCRF). Combining CRF with nCRF could increase the sparseness, reliability and precision of the neuronal responses. This paper is aimed at the realization of the neural circuit and the dynamic adjustment mechanism of the receptive field (RF) with respect to nCRF. On the basis of anatomical and electrophysiological evidence, we constructed a neural computational model, which can represent natural images faithfully, simply and rapidly. And the representation can significantly improve the subsequent operation efficiency such as segmentation or integration. This study is of particular significance in the development of efficient image processing algorithms based on neurobiological mechanisms. The RF mechanism of ganglion cell (GC) is the result of a long term of evolution and optimization of self-adaptability and high representation efficiency. So its performance evaluation in natural image processing is worthy of further study.

关键词： Radio frequency Retina Computational modeling Integrated circuit modeling Mathematical model Equations Feedback control

来源：评论

学校读者我要写书评

暂无评论

A structural representing and learning model based on biological neural mechanism

A structural representing and learning model based on biolog...

引用

International Conference on Machine Learning and Cybernetics (ICMLC)

作者： Hui Wei Hui-Xuan Tang Department of Computer Science Laboratory of Algorithm for Cognitive Model Brain Science Research Center Fudan University Shanghai China

ISBN: (纸本)0780384032

From the view of cognitive computational neuroscience, a direct representing method based on neural dynamics and graph theory is presented. Firstly, an assembly of neuron as well as its dynamics is defined. They directly represent the perceptual information of stimulus. Then a two layer neural network is designed to retain the features of that stimulus and generate the very neural circuit responding to it. This is achieved by the structure-learning algorithm. The circuit can also serve as an associative base whose credibility is decided by its connectivity. The direct representing method is of great significance in the research of semantic representation and semantic-driven inference in artificial intelligence as well as in artificial neural network researches. The exhibition of major physiological features of neural information processing distinguishes this model from the traditional ones.

关键词： Biological system modeling Circuits Biology computing Neuroscience Graph theory Assembly Neurons Biological neural networks Inference algorithms Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Object Topological Character Acquisition by Inductive Learning

arXiv

引用

arXiv 2023年

作者： Hui, Wei Yu, Liping Wei, Yiran Laboratory of Algorithm for Cognitive Model School of Computer Science Shanghai Key Laboratory of Data Science Fudan University China Shanghai Pinghe School China

Understanding the shape and structure of objects is undoubtedly extremely important for object recognition, but the most common pattern recognition method currently used is machine learning, which often requires a large number of training data. The problem is that this kind of object-oriented learning lacks a priori knowledge. The amount of training data and the complexity of computations are very large, and it is hard to extract explicit knowledge after learning. This is typically called ‘knowing how without knowing why’. We adopted a method of inductive learning, hoping to derive conceptual knowledge of the shape of an object and its formal representation based on a small number of positive examples. It is clear that implementing object recognition is not based on simple physical features such as colors, edges, textures, etc., but on their common geometry, such as topologies, which are stable, persistent, and essential to recognition. In this paper, a formal representation of topological structure based on object’s skeleton (RTS) was proposed and the induction process of’seeking common ground’ is realized. This research helps promote the method of object recognition from empiricism to rationalism © 2023, CC BY-NC-ND.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：