Background neural networks with uniform firing rate and background input

Huajin Tang,Kaychen Tan,Zhang Yi

DOI: https://doi.org/10.1007/978-3-540-69226-3_17

2007-01-01

Abstract:null

What problem does this paper attempt to address?

Dynamical Properties of Continuous Attractor Neural Network with Background Tuning

Jiali Yu,Huajin Tang,Haizhou Li,Luping Shi

DOI: https://doi.org/10.1016/j.neucom.2012.06.029

IF: 6

2013-01-01

Neurocomputing

Abstract:Persistent activity holds the transient stimulus for up to many seconds even after the stimulus is gone. It has been implemented in a class of models known as continuous attractor neural networks, which have infinite stable states corresponding to persistent activity patterns. Continuous attractor neural network remains stable so does not change systematically in the absence of stimulus input. Continuous attractor is a set of connected stable equilibrium points and has been used to describe the storing of continuous stimuli in neural networks. The background input of the networks plays an important role in continuous attractor neural network. In this paper, dynamical properties of continuous attractor neural network with two background input tuning schemes are investigated: constant input shifting and oscillation background activity. Simulations are employed to illustrate the theory.
Feedforward neural networks and training methods

Huajin Tang,Kaychen Tan,Zhang Yi

DOI: https://doi.org/10.1007/978-3-540-69226-3_2

2007-01-01

Abstract:null
Background Subtraction Based on Nonparametric Bayesian Estimation.

Yan He,Donghui Wang,Miaoliang Zhu

DOI: https://doi.org/10.1117/12.896509

2011-01-01

Abstract:Background subtraction, the task of separating foreground pixels from background pixels in a video, is an important step in video processing. Comparing with the parametric background modeling methods, nonparametric methods use a model selection criterion to choose the right number of components for each pixel online. We model the background subtraction problem with the Dirichlet process mixture, which constantly adapts both the parameters and the number of components of the mixture to the scene.
Background initialization algorithm in traffic flow video detection

Li Zhi-hui,Zhang Chang-hai,Qu Zhao-wei,Wei Wei,Dianhai Wang

DOI: https://doi.org/10.13229/j.cnki.jdxbgxb2008.01.037

2008-01-01

Abstract:To get the initialized background from the training sequence with foreground objects,a new background initialization algorithm was proposed based on clustering classifier.In this method,all stable sub-intervals in the training sequence were located for each pixel as possible background.Then a classify data set was constructed by the median values of each sub-interval.A background sub-set was obtained from the data set by unsupervised clustering.Accordingly the initialized background was obtained from the background subset.By experiments,the proposed method was compared with the median method and the sample consensus method under different traffic conditions.Results show that this method is robust,overcomes the influence of slow moving big objects,and tolerates over 50% foreground pixels and noises.
Competitive neural networks for image segmentation

Huajin Tang,Kaychen Tan,Zhang Yi

DOI: https://doi.org/10.1007/978-3-540-69226-3_9

2007-01-01

Abstract:null
Convergence analysis of background neural networks with two subnetworks

Fang Xu,Lei Zhang,Hong Qu

DOI: https://doi.org/10.1109/ICCIS.2008.4670913

2008-01-01

Abstract:The convergence of background neural networks with two identical, unstructured subnetworks that inhibit each other is studied. The convergence properties include global attractivity and complete stability analysis. Furthermore, it shows that shifting the background level affects the existence and stability of the equilibrium point. Adding noise to the networks, it will fluctuate around an equilibrium point.
Time for a Background Check! Uncovering the impact of Background Features on Deep Neural Networks

Vikash Sehwag,Rajvardhan Oak,Mung Chiang,Prateek Mittal

DOI: https://doi.org/10.48550/arXiv.2006.14077

2020-06-25

Abstract:With increasing expressive power, deep neural networks have significantly improved the state-of-the-art on image classification datasets, such as ImageNet. In this paper, we investigate to what extent the increasing performance of deep neural networks is impacted by background features? In particular, we focus on background invariance, i.e., accuracy unaffected by switching background features and background influence, i.e., predictive power of background features itself when foreground is masked. We perform experiments with 32 different neural networks ranging from small-size networks to large-scale networks trained with up to one Billion images. Our investigations reveal that increasing expressive power of DNNs leads to higher influence of background features, while simultaneously, increases their ability to make the correct prediction when background features are removed or replaced with a randomly selected texture-based background.

Computer Vision and Pattern Recognition,Machine Learning
BGnet: Accurate and rapid background estimation in single-molecule localization microscopy with deep neural nets

Leonhard Möckl,Anish R. Roy,Petar N. Petrov,W.E. Moerner

DOI: https://doi.org/10.1073/pnas.1916219117

2019-09-18

Abstract:Background fluorescence, especially when it exhibits undesired spatial features, is a primary factor for reduced image quality in optical microscopy. Structured background is particularly detrimental when analyzing single-molecule images for 3D localization microscopy or single-molecule tracking. Here, we introduce BGnet, a deep neural network with a U-net-type architecture, as a general method to rapidly estimate the background underlying the image of a point source with excellent accuracy, even when point spread function (PSF) engineering is in use to create complex PSF shapes. We trained BGnet to extract the background from images of various PSFs and show that the identification is accurate for a wide range of different interfering background structures constructed from many spatial frequencies. Furthermore, we demonstrate that the obtained background-corrected PSF images, both for simulated and experimental data, lead to a substantial improvement in localization precision. Finally, we verify that structured background estimation with BGnet results in higher quality of super-resolution reconstructions of biological structures.

Optics,Applied Physics
Inl: Implicit Non-Local Network

Yifeng Han,Xi Chen,Songjie Zhang,Donglian Qi

DOI: https://doi.org/10.1016/j.neucom.2022.01.047

IF: 6

2022-01-01

Neurocomputing

Abstract:The attention mechanism of computer vision represented by a non-local network improves the performance of numerous vision tasks while bringing computational burden for deployment Wang et al. (2018). In this work, we explore to release the inference computation for non-local network by decoupling the training/inference procedure. Specifically, we propose the implicit non-local network (iNL). During training, iNL models the dependency between features across long-range affinities like original non-local blocks; during inference, iNL could be reformulated as only two convolution layers but can rival non-local network. In this way, the computation complexity and the memory costs are reduced. In addition, we take a further step and extend our iNL into a more generalized form, which covers the attentions of different orders in computer vision tasks. iNL brings steady improvements on multiple benchmarks of different vision tasks including classification, detection, and instance segmentation. In the meantime, it provides a brand–new perspective to understand the attention mechanism in deep neural networks.
Reduction of Class Activation Uncertainty with Background Information

H M Dipu Kabir

2024-07-14

Abstract:Multitask learning is a popular approach to training high-performing neural networks with improved generalization. In this paper, we propose a background class to achieve improved generalization at a lower computation compared to multitask learning to help researchers and organizations with limited computation power. We also present a methodology for selecting background images and discuss potential future improvements. We apply our approach to several datasets and achieve improved generalization with much lower computation. Through the class activation mappings (CAMs) of the trained models, we observed the tendency towards looking at a bigger picture with the proposed model training methodology. Applying the vision transformer with the proposed background class, we receive state-of-the-art (SOTA) performance on STL-10, Caltech-101, and CINIC-10 datasets. Example scripts are available in the 'CAM' folder of the following GitHub Repository: <a class="link-external link-http" href="http://github.com/dipuk0506/UQ" rel="external noopener nofollow">this http URL</a>

Computer Vision and Pattern Recognition
Deep Background Modeling Using Fully Convolutional Network.

Lu Yang,Jing Li,Yuansheng Luo,Yang Zhao,Hong Cheng,Jun Li

DOI: https://doi.org/10.1109/TITS.2017.2754099

IF: 8.5

2018-01-01

IEEE Transactions on Intelligent Transportation Systems

Abstract:Background modeling plays an important role for video surveillance, object tracking, and object counting. In this paper, we propose a novel deep background modeling approach utilizing fully convolutional network. In the network block constructing the deep background model, three atrous convolution branches with different dilate are used to extract spatial information from different neighborhoods o...
Faster ISNet for Background Bias Mitigation on Deep Neural Networks

Pedro R. A. S. Bassi,Sergio Decherchi,Andrea Cavalli

2024-04-01

Abstract:Bias or spurious correlations in image backgrounds can impact neural networks, causing shortcut learning (Clever Hans Effect) and hampering generalization to real-world data. ISNet, a recently introduced architecture, proposed the optimization of Layer-Wise Relevance Propagation (LRP, an explanation technique) heatmaps, to mitigate the influence of backgrounds on deep classifiers. However, ISNet's training time scales linearly with the number of classes in an application. Here, we propose reformulated architectures whose training time becomes independent from this number. Additionally, we introduce a concise and model-agnostic LRP implementation. We challenge the proposed architectures using synthetic background bias, and COVID-19 detection in chest X-rays, an application that commonly presents background bias. The networks hindered background attention and shortcut learning, surpassing multiple state-of-the-art models on out-of-distribution test datasets. Representing a potentially massive training speed improvement over ISNet, the proposed architectures introduce LRP optimization into a gamut of applications that the original model cannot feasibly handle.

Image and Video Processing,Computer Vision and Pattern Recognition,Computers and Society,Machine Learning
Clustering and Observation on Neuron Tuning Property for Brain Machine Interfaces

Xiwei She,Yuxi Liao,Hongbao Li,Qiaosheng Zhang,Yiwen Wang,Xiaoxiang Zheng

DOI: https://doi.org/10.1109/mfi.2014.6997680

2014-01-01

Abstract:Neurons' tuning describes how the neural activity responses to the stimulus. As the prior knowledge, understanding more about the neuron tuning helps better decode the movement information from the neural firings for brain machine interfaces. We are interested in qualifying the neural tuning and observe whether there are similar tunings among the ensemble recordings and how they change over time. We propose to implement a linear-nonlinear-Poisson model to describe the neural tuning function. And the function parameters build a feature space, where the neuron tuning characters can be visually observed. We use k-means algorithm to cluster neuron tuning characters and find that there are three types of neurons with different tuning curve shapes. The nonlinear-shaping neurons are not majority in number but have important contribution (evaluated by mutual information) relative to the movement task than the linear ones. Furthermore, we find some neuron tunings shows clear time-varying properties in the feature space, which can be predicted by a random walk model. And we prove it through two kinds of way: Kernel size-CC estimation and Kolmogorov-Smirnov plot (KS plot). The predictable time-varying tuning suggests a better understanding of neuron property and potentially contributes to decode the non-stationary neuron activities.
Neural Networks: Computational Models and Applications - ReadingSample

Huajin Tang,K. Tan,Zhang Yi

2015-01-01

Abstract:1 Introduction Typically, models of neural networks are divided into two categories in terms of signal transmission manner: feed-forward neural networks and recurrent neural networks. They are built up using different frameworks, which give rise to different fields of applications. Feed-forward neural network (FNN), also referred to as multilayer percep-trons (MLPs), has drawn great interests over the last two decades for its distinction as a universal function approximator As an important intelligent computation method, FNN has been applied to a wide range of applications, including curve fitting, pattern classification and nonlinear system identification and so on (Vemuri, 1995). FNN features a supervised training with a highly popular algorithm known as the error back-propagation algorithm. In the standard back-propagation (SBP) algorithm, the learning of a FNN is composed of two passes: in the forward pass, the input signal propagates through the network in a forward direction, on a layer-by-layer basis with the weights fixed; in the backward pass, the error signal is propagated in a backward manner. The weights are adjusted based on an error-correction rule. Although it has been successfully used in many real world applications, SBP suffers from two infamous shortcomings, i.e., slow learning speed and sensitivity to parameters. Many iterations are required to train small networks, even for a simple problem. The sensitivity to learning parameters, initial states and perturbations was analyzed in (Yeung and Sun, 2002). Behind such drawbacks the learning rate plays a key role in affecting the learning performance and it has to be chosen carefully. If the learning rate is large, the network may exhibit chaotic
Nonparametric Background Generation

Yazhou Liu,Hongxun Yao,Wen Gao,Xilin Chen,Debin Zhao

DOI: https://doi.org/10.1016/j.jvcir.2007.01.003

IF: 2.887

2007-01-01

Journal of Visual Communication and Image Representation

Abstract:A novel background generation method based on non-parametric background model is presented for background subtraction. We introduce a new model, named as effect components description (ECD), to model the variation of the background, by which we can relate the best estimate of the background to the modes (local maxima) of the underlying distribution. Based on ECD, an effective background generation method, most reliable background mode (MRBM), is developed. The basic computational module of the method is an old pattern recognition procedure, the mean shift, which can be used recursively to find the nearest stationary point of the underlying density function. The advantages of this method are three fold: first, backgrounds can be generated from image sequence with cluttered moving objects; second, backgrounds are very clear without blur effect; third, it is robust to noise and small vibration. Extensive experimental results illustrate its good performance.
What Happens Without Background? Constructing Foreground-Only Data for Fine-Grained Tasks

Yuetian Wang,Wenjin Hou,Qinmu Peng,Xinge You

2024-08-04

Abstract:Fine-grained recognition, a pivotal task in visual signal processing, aims to distinguish between similar subclasses based on discriminative information present in samples. However, prevailing methods often erroneously focus on background areas, neglecting the capture of genuinely effective discriminative information from the subject, thus impeding practical application. To facilitate research into the impact of background noise on models and enhance their ability to concentrate on the subject's discriminative features, we propose an engineered pipeline that leverages the capabilities of SAM and Detic to create fine-grained datasets with only foreground subjects, devoid of background. Extensive cross-experiments validate this approach as a preprocessing step prior to training, enhancing algorithmic performance and holding potential for further modal expansion of the data.

Computer Vision and Pattern Recognition
Application Of ADNN For Background Subtraction In Smart Surveillance System

Piyush Batra,Gagan Raj Singh,Neeraj Goyal

DOI: https://doi.org/10.48550/arXiv.2301.00264

2023-01-01

Abstract:Object movement identification is one of the most researched problems in the field of computer vision. In this task, we try to classify a pixel as foreground or background. Even though numerous traditional machine learning and deep learning methods already exist for this problem, the two major issues with most of them are the need for large amounts of ground truth data and their inferior performance on unseen videos. Since every pixel of every frame has to be labeled, acquiring large amounts of data for these techniques gets rather expensive. Recently, Zhao et al. [1] proposed one of a kind Arithmetic Distribution Neural Network (ADNN) for universal background subtraction which utilizes probability information from the histogram of temporal pixels and achieves promising results. Building onto this work, we developed an intelligent video surveillance system that uses ADNN architecture for motion detection, trims the video with parts only containing motion, and performs anomaly detection on the trimmed video.

Computer Vision and Pattern Recognition,Artificial Intelligence
Adaptive Background Model Based on Hybrid Structure Neural Network

WANG Zhi-ming,ZHANG Li,BA Hong

2011-01-01

Abstract:This paper proposed a new background model for motion detection in video surveillance based on neural network(NN).A neural network background model was build for every pixel(or a small local region).It is a four-layer feedforward neural network.Input layer accept HSV pixel value,feature layer extract features form HSV,pattern layer work as a background probability calculator.Output layer classifies the pixel into background or foreground,and finds the activated node.Weights and structure of network updated dynamically along with motion detection and no training video needed.Adaptability of background model includes adaptive learning rate calculated form motion difference between adjacent frames,and number of pattern node changes according to weight variation.Experimental results on benchmark videos show that,without any manual setting of learning rate,the proposed algorithm can detection motion more precisely than other familiar background models,and it can also adapts to sudden background or lighting changes more quickly.
Dynamic Background Learning Through Deep Auto-Encoder Networks

Pei Xu,Mao Ye,Xue Li,Qihe Liu,Yi Yang,Jian Ding

DOI: https://doi.org/10.1145/2647868.2654914

2014-01-01

Abstract:Background learning is a pre-processing of motion detection which is a basis step of video analysis. For the static background, many previous works have already achieved good performance. However, the results on learning dynamic background are still much to be improved. To address this challenge, in this paper, a novel and practical method is proposed based on deep auto-encoder networks. Firstly, dynamic background images are extracted through a deep auto-encoder network (called Background Extraction Network) from video frames containing motion objects. Then, a dynamic background model is learned by another deep auto-encoder network (called Background Learning Network) using the extracted background images as the input. To be more flexible, our background model can be updated on-line to absorb more training samples. Our main contributions are 1) a cascade of two deep auto-encoder networks which can deal with the separation of dynamic background and foregrounds very efficiently; 2) a method of online learning is adopted to accelerate the training of Background Extraction Network. Compared with previous algorithms, our approach obtains the best performance over six benchmark data sets. Especially, the experiments show that our algorithm can handle large variation background very well.
Where-What Network With Cuda: General Object Recognition And Location In Complex Backgrounds

Yuekai Wang,Xiaofeng Wu,Xiaoying Song,Wengqiang Zhang,Juyang Weng

DOI: https://doi.org/10.1007/978-3-642-21090-7_39

2011-01-01

Abstract:An effective framework for general object recognition and localization from complex backgrounds had not been found till the brain-inspired Where-What Network (WWN) series by Weng and coworkers. This paper reports two advances along this line. One is the automatic adaptation of the receptive field of each neuron to disregard input dimensions that arise from backgrounds but without a handcrafted object model, since the initial hexagonal receptive field does not tit well the contour of the automatically assigned object view. The other is the hierarchical parallelization technique and its implementation on the GPU-based accelerator using the CUDA parallel language. The experimental results showed that automatic adaptation of the receptive fields led to improvements in the recognition rate. The hierarchical parallelization technique has achieved a speedup of 16 times compared to the C program. This speed-up was employed on the Haibao Robot displayed at the World Expo, Shanghai 2010.

Background neural networks with uniform firing rate and background input

Dynamical Properties of Continuous Attractor Neural Network with Background Tuning

Feedforward neural networks and training methods

Background Subtraction Based on Nonparametric Bayesian Estimation.

Background initialization algorithm in traffic flow video detection

Competitive neural networks for image segmentation

Convergence analysis of background neural networks with two subnetworks

Time for a Background Check! Uncovering the impact of Background Features on Deep Neural Networks

BGnet: Accurate and rapid background estimation in single-molecule localization microscopy with deep neural nets

Inl: Implicit Non-Local Network

Reduction of Class Activation Uncertainty with Background Information

Deep Background Modeling Using Fully Convolutional Network.

Faster ISNet for Background Bias Mitigation on Deep Neural Networks

Clustering and Observation on Neuron Tuning Property for Brain Machine Interfaces

Neural Networks: Computational Models and Applications - ReadingSample

Nonparametric Background Generation

What Happens Without Background? Constructing Foreground-Only Data for Fine-Grained Tasks

Application Of ADNN For Background Subtraction In Smart Surveillance System

Adaptive Background Model Based on Hybrid Structure Neural Network

Dynamic Background Learning Through Deep Auto-Encoder Networks

Where-What Network With Cuda: General Object Recognition And Location In Complex Backgrounds