Olshausen and Field (1996) developed a simple cell receptive field model for natural scene processing in V1, based on unsupervised learning and non-orthogonal basis function optimization of an overcomplete representat...
详细信息
Network coding can ease the block scheduling and thus makes the distribution more efficient. However, the complexities of encoding and decoding increase sharply as the content size scales up. In this paper, we propose...
详细信息
ISBN:
(纸本)9781424450053
Network coding can ease the block scheduling and thus makes the distribution more efficient. However, the complexities of encoding and decoding increase sharply as the content size scales up. In this paper, we propose a coding scheme which combines chunked coding and sparse linear network coding to reduce both encoding and decoding costs of content distribution. Based on the combined scheme, we implement a P2P content distribution system, named Dasher, where Local-Rarest-First scheme is adopted for chunk scheduling. Under the same system architecture, we implement three comparative systems, a BitTorrent-like system, named Mybt, a system only with sparse coding, named sparser and a system using chunked coding, named Chunker. We conduct extensive experiments to compare the performance among these four systems. The experimental results show that Dasher with certain chunk sizes can reduce the average downloading time up to 15% compared with Mybt, and up to 43% with Chunker. With proper chunk sizes, the downloading time of Dasher is almost the same with sparser. The average decoding rate of Dasher is the same with Chunker, and is nearly m times as fast as sparser, where m is the number of chunks. Moreover, with respect to robustness, Dasher performs almost as well as Chunker, better than Mybt, but worse than sparser.
O-glycosylation is one of the main types of the mammalian protein glycosylation, it occurs on the particular site of serine and threonine. In this paper, a new method of PCA-LDA is used for the prediction of O-glycosy...
详细信息
O-glycosylation is one of the main types of the mammalian protein glycosylation, it occurs on the particular site of serine and threonine. In this paper, a new method of PCA-LDA is used for the prediction of O-glycosylation site under all kinds of window size (5,7,9,11,21,31,41,51). The new method of PCA-LDA is the combination of PCA and LDA, we also call it hybrid discriminant analysis(HDA). The test protein sequence which is encoded by the sparse coding is projected to the one-dimensional subspace and then by calculating the Mahanalobis distance between the projection and each class center, the test protein sequence is assigned into the “nearest” class, so it can be known that whether a particular site of serine and threonine is glycosylated. The result of experiments shows that the proposed method of HDA is more effective and accurate. The prediction accuracy is about 75%-92.5%.
Image representation has been a key issue in vision research for many years. In order to represent various local image patterns or objects effectively, it is important to study the spatial relationship among these obj...
详细信息
Image representation has been a key issue in vision research for many years. In order to represent various local image patterns or objects effectively, it is important to study the spatial relationship among these objects, especially for the purpose of searching the specific object among them. Psychological experiments have supported the hypothesis that humans cognize the world using visual context or object spatial relationship. How to efficiently learn and memorize such knowledge is a key issue that should be studied. This paper proposes a new type of neural network for learning and memorizing object spatial relationship by means of sparse coding. A group of comparison experiments for visual object searching between several sparse features are carried out to examine the proposed approach. The efficiency of sparse coding of the spatial relationship is analyzed and discussed. Theoretical and experimental results indicate that the newly developed neural network can well learn and memorize object spatial relationship and simultaneously the visual context learning and memorizing have certainly become a grand challenge in simulating the human vision system. (C) 2008 Elsevier B.V. All rights reserved.
We consider the problem of convolutive blind source separation of stereo mixtures, where a pair of microphones records mixtures of sound sources that are convolved with the impulse response between each source and sen...
详细信息
We consider the problem of convolutive blind source separation of stereo mixtures, where a pair of microphones records mixtures of sound sources that are convolved with the impulse response between each source and sensor. We propose an adaptive stereo basis (ASB) source separation method for such convolutive mixtures, using an adaptive transform basis which is learned from the stereo mixture pair. The stereo basis vector pairs of the transform are grouped according to the estimated relative delay between the left and right channels for each basis, and the sources are then extracted by projecting the transformed signal onto the subspace corresponding to each group of basis vector pairs. The performance of the proposed algorithm is compared with FD-ICA and DUET under different reverberation and noise conditions, using both objective distortion measures and formal listening tests. The results indicate that the proposed stereo coding method is competitive with both these algorithms at short and intermediate reverberation times, and offers significantly improved performance at low noise and short reverberation times. (C) 2008 Elsevier B.V. All rights reserved.
Philosophers have pointed out that there is a close relation between the esthetics of art and the beauty of natural scenes. Supporting this similarity at the experimental level, we have recently shown that visual art ...
详细信息
Philosophers have pointed out that there is a close relation between the esthetics of art and the beauty of natural scenes. Supporting this similarity at the experimental level, we have recently shown that visual art and natural scenes share fractal-like, scale-invariant statistical properties. Moreover, evidence from neurophysiological experiments shows that the visual system uses an efficient (sparse) code to process optimally the statistical properties of natural stimuli. In the present work, a hypothetical model of esthetic perception is described that combines both lines of evidence. Specifically, it is proposed that an artist creates a work of art so that it induces a specific resonant state in the visual system. This resonant state is thought to be based on the adaptation of the visual system to natural scenes. The proposed model is universal and predicts that all human beings share the same general concept of esthetic judgment. The model implies that esthetic perception, like the coding of natural stimuli, depends on stimulus form rather than content, depends on higher-order statistics of the stimuli, and is non-intuitive to cognitive introspection. The model accommodates the central tenet of neuroesthetic theory that esthetic perception reflects fundamental functional properties of the nervous system.
This paper presents learning multilayer Potts perceptrons (MLPotts) for data driven function approximation. A Potts perceptron is composed of a receptive field and a K-state transfer function that is generalized from ...
详细信息
This paper presents learning multilayer Potts perceptrons (MLPotts) for data driven function approximation. A Potts perceptron is composed of a receptive field and a K-state transfer function that is generalized from sigmoid-like transfer functions of traditional perceptrons. An MLPotts network is organized to perform translation from a high-dimensional input to the sum of multiple postnonlinear projections, each with its own postnonlinearity realized by a weighted K-state transfer function. MLPotts networks span a function space that theoretically covers network functions of multilayer perceptrons. Compared with traditional perceptrons, weighted Potts perceptrons realize more flexible postnonlinear functions for nonlinear mappings. Numerical simulations show MLPotts learning by the Levenberg-Marquardt (LM) method significantly improves traditional supervised learning of multilayer perceptrons for data driven function approximation.
Evidence that neurosensory systems use sparse signal representations as well as improved performance of signal processing algorithms using sparse signal models raised interest in sparse signal codingin the last years....
详细信息
Evidence that neurosensory systems use sparse signal representations as well as improved performance of signal processing algorithms using sparse signal models raised interest in sparse signal codingin the last years. For natural. audio signals like speech and environmental sounds, gammatone atoms have been derived as expansion functions that generate a nearly optimal sparse signal model (Smith, E., Lewicki, M., 2006. Efficient auditory coding. Nature 439, 978-982). Furthermore, gammatone functions are established models for the human auditory filters. Thus far, a practical application of a sparse gammatone signal model has been prevented by the fact that deriving the sparsest representation is, in general, computationally intractable. In this paper, we applied an accelerated version of the matching pursuit algorithm for gammatone dictionaries allowing real-time and large data set applications. We show that a sparse signal model in general has advantages in audio coding and that a sparse gammatone signal model encodes speech more efficiently in terms of sparseness than a sparse modified discrete cosine transform (MDCT) signal model. We also show that the optimal gammatone parameters derived for English speech do not match the human auditory filters, suggesting for signal processing applications to derive the parameters individually for each applied signal class instead of using psychometrically derived parameters. For brain research, it means that care should be taken with directly transferring findings of optimality for technical to biological systems. (C) 2007 Elsevier B.V. All rights reserved.
There is no reasonable doubt that the hippocampus plays an important role in memory processing. A virtually uncountable number of studies in animals and humans have revealed changes in neural activity in this structur...
详细信息
There is no reasonable doubt that the hippocampus plays an important role in memory processing. A virtually uncountable number of studies in animals and humans have revealed changes in neural activity in this structure during memory formation [Squire LR. Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans. Psychol Rev 1992;99:195-231;Squire LR, Stark CE, Clark RE. The medial temporal lobe. Annu Rev Neurosci 2004;27:279-306], and hippocampal damage reliably leads to impairments in a large number of memory tests. However, while several correlates of successful memory formation have been found in the hippocampus, it is still an open question why specific neural processes support encoding of a particular item. An answer to this question would help to resolve current debates about which memory functions are actually supported by the hippocampus, and why activity in the neural networks of the hippocampus is involved in, or even necessary for, some memory processes but not for others. In this review, we first summarize findings on the electrophysiological activity within the hippocampus during different memory processes. We try to differentiate whether the hippocampus is merely involved in these processes, or whether the hippocampus appears to be necessary for them. Based on a distinction between a more general "encoding state" and the more specific process of "content-specific memory formation", we review data on neural representations within hippocampus and neocortex. We suggest that during memory formation, the hippocampus renders neural representations more sparse by providing an inhibitory signal to the neocortex. (c) 2007 Elsevier B.V. All rights reserved.
We propose a sparse non-negative image coding based on simulated annealing and matrix pseudo-inversion. We show that sparsity and non-negativity are both important to obtain part-based coding and we also show the impa...
详细信息
ISBN:
(纸本)9781424414833
We propose a sparse non-negative image coding based on simulated annealing and matrix pseudo-inversion. We show that sparsity and non-negativity are both important to obtain part-based coding and we also show the impact of each of them on the coding. In contrast with other approaches in the literature, our method can constrain both weights and basis vectors to generate part-based bases suitable for image recognition and fiducial point extraction. We also propose a speed-up of the algorithm by implementing a hybrid system that mixes simulated annealing and pseudo-inverse computation of matrices.
暂无评论