Showing papers on "Encoding (memory) published in 1989"

Implicit Memory: Theoretical Issues

Abstract: Contents: Part I:Introduction. R.S. Lockhart, The Role of Theory in Understanding Implicit Memory. Part II:Characterizing Implicit Memory. J.C. Dunn, K. Kirsner, Implicit Memory: Task or Process? W. Hirst, On Consciousness, Recall, Recognition, and the Architecture of Memory. D.L. Schacter, J. Bowers, J. Booker, Intention, Awareness, and Implicit Memory: The Retrieval Intentionality Criterion. H.L. Roediger, K. Srinivas, M.S. Weldon, Dissociations Between Implicit Measures of Retention. Part III:Theories and Models. E.U. Weber, B.B. Murdock, Priming in a Distributed Memory System: Implications for Models of Implicit Memory. K. Kirsner, J.C. Dunn, P. Standen, Domain-Specific Resources in Word Recognition. M.E.J. Masson, Fluent Preprocessing as an Implicit Expression of Memory for Experience. M.S. Humphreys, J.D. Bain, J.S. Burt, Episodically Unique and Generalized Memories: Applications to Human and Animal Amnesics. Part IV:Processes and Representations. C.M. MacLeod, J.N. Bassili, Are Implicit and Explicit Tests Differentially Sensitive to Item-Specific vs. Associative Information? L-G. Nilsson, L. B ckman, Implicit Memory and the Enactment of Verbal Instructions. S. Lewandowsky, K. Kirsner, V. Bainbridge, Context Effects in Implicit Memory: A Sense Specific Account. M. Carroll, Implicit Memory: Compatibility Between Study-Test Operations. S. Kinoshita, Masked and Unmasked Repetition Effects: Activation of Representation or Procedure? Part V:Development and Learning. A.J. Parkin, The Development and Nature of Implicit Memory. K. Durkin, Implicit Memory and Language Acquisition. J.G. Snodgrass, Sources of Learning in the Picture Fragment Completion Task. Part VI:Comment. M. Coltheart, Implicit Memory and the Functional Architecture of Cognition.

A fast algorithm for computing distance spectrum of convolutional codes

Abstract: A fast algorithm for searching a tree (FAST) is presented for computing the distance spectrum of convolutional codes. The distance profile of a code is used to limit substantially the error patterns that have to be searched. The algorithm can easily be modified to determine the number of nonzero information bits of an incorrect path as well as the length of an error event. For testing systematic codes, a faster version of the algorithm is given. FAST is much faster than the standard bidirectional search. On a microVAX, d/sub infinity /=27 was verified for a rate R=1/2, memory M=25 code in 37 s of CPU time. Extensive tables of rate R=1/2 encoders are given. Several of the listed encoders have distance spectra superior to those of any previously known codes of the same rate and memory. A conjecture than an R=1/2 systematic convolutional code of memory 2M will perform as well as a nonsystematic convolutional code of memory M is given strong support. >

Correlational analyses of explicit and implicit memory performance

Abstract: Sixty-four subjects were administered two tests of explicit memory (selective recall and recognition) and four tests of implicit memory (identification in a perceptual clarification procedure, word-fragment completion, tachistoscopic identification, and anagram solution). Each test drew on a different subset of a long list of previously displayed words. Although the four implicit memory tests showed sizable priming effects, correlational and factor analyses showed striking dissociations. On the one hand, performance on the perceptual clarification procedure and word-completion tests were related to one another, as well as to recall and recognition. On the other hand, performance on tachistoscopic identification and anagram solution were related to one another, but not to the measures for the other tasks. A framework is proposed to reconcile these new results with current knowledge on the explicit/implicit memory distinction, based in particular on studies of amnesic subjects. It is argued that a small number of tasks, especially tachistoscopic identification, may serve as relatively uncontaminated and ubiquitous indicators of implicit memory. However, explicit remembering could affect performance in so-called implicit memory tasks that allow for a strategy of controlled selection of candidate responses from accumulating cues, in experimental conditions that make the explicit remembering of relevant events possible.

Performance analysis and application of the bidirectional associative memory to industrial spectral signatures

Abstract: A detailed study of the performance of the bidirectional associative memory (BAM) for the classification of power spectral density (PSD) functions is presented. The BAM has error-correcting properties and is capable of generating a successful recall, even with noisy and incomplete patterns. Suitable pattern coding schemes are utilized to enhance the discriminating features in the patterns. A bit-mapping scheme is used to construct binary pattern vectors from the PSD. The numerical aspects of signature encoding and fault tolerance property of the BAM are evaluated using spectral density functions of signals from a fiber-manufacturing rotating machinery system. The results of the application of the BAM are superior to those obtained with classical pattern recognition methods. >

Hierarchical vector quantisation

Abstract: We present a method of vector quantisation which trades off accuracy for speed of encoding. We achieve this by hierarchically structuring a multistage encoder so that each stage encodes low dimensional input vectors. Such hierarchical encoders may easily be realised as a set of fast table look-up operations. We demonstrate how the Euclidean distortion in such a multistage encoder is approximately minimised by using Kohonen's topographic mapping learning algorithm from neural network theory. We also demonstrate the performance of the technique on various stochastic time series. We find that there is little loss in encoding accuracy, when compared with the exact nearest neighbour encoding using an equivalent single stage encoder.

Short-Term Memory for Associations

Abstract: Publisher Summary This chapter discusses short-term memory for associations The experiments described in the chapter test the recognition of associative information over retention intervals filled with the presentation and testing of other paired associates The recognition is used rather than recall because the recognition provides a less complex view of associative information than the recall A decision process occurs in recognition that may be absent in recall Signal-detection methods applied to memory allows to separate strength effects from criterion effects in recognition memory These experiments are designed to trace out the retention curve for the recognition of associative information using a continuous paired-associates paradigm The convolution–correlation model is a distributed-memory model that assumes that all information is stored in a common memory Items are represented by random vectors and the numerical value of N is one of the parameters of the model Convolution is the associative operation and correlation is the retrieval operation

On write-unidirectional memory codes

Abstract: Write-unidirectional memories generalize write-once memories storing binary sequences of some fixed length in a reusable manner. At every new usage the content of the memory can be rewritten by either changing some of the zeroes to ones or changing some of the ones to zeroes, but not both. The author constructs codes of rate 0.5325. He discusses the four cases that arise according to whether or not the encoder and/or the decoder is informed of the previous state of the memory. J.M. Borden's converse bound (submitted to IEEE Trans. Inf. Theory) is rederived using Fibonacci sequences. >

High-order bidirectional associative memory

Abstract: A novel encoding scheme of a bidirectional associative memory (BAM) incorporating the high-order nonlinearity is proposed. This method significantly improves the storage capacity and error-correcting capability of the BAM.

Performance of back propagation networks for associative database retrieval

Abstract: Back-propagation networks have been successfully used to perform a variety of input-output mapping tasks for recognition, generalization, and classification. In spite of this method's popularity, virtually nothing is known about its saturation/capacity and, in more general terms, about its performance as an associative memory. The authors address these issues using associative database retrieval as an original application domain. Experimental results show that the quality of recall and the network capacity are very significantly affected by the network topology (the number of hidden units), data representation (encoding), and the choice of learning parameters. On the basis of their results and the fact that back-propagation learning is not recursive, the authors conclude that back-propagation networks can be used mainly as read-only associative memories and represent a poor choice for read-and-write associative memories. >

Why does imagery improve memory

Abstract: Evidence over the last 15 years has suggested that dual (imagery and verbal) coding explanations of concreteness effects in memory for word lists do not generalise well to memory for sentences and paragraphs. In contrast, an alternative framework based on relative differences in relational and distinctive processing has been shown to account for the effects of imagery and concreteness in these contexts and others. This paper describes recent research on free and cued recall of word lists and evaluates it with respect to the two models. The evidence suggests that whereas dual processing systems may be involved in the encoding of verbal materials, dual memory codes are insufficient to explain concreteness effects in recall. Better memory for high-as compared to low-imagery words depends on the use of paradigms that facilitate inter-item relational processing, independent of whether or not imagery is involved.

Associative polarization-encoded optical shadow casting: gray-level image encoding for serial and parallel operations

Abstract: An efficient algorithm for designing a serial and parallel multioutput logic unit using an associative polarization-encoded optical shadow-casting technique is presented. Devices designed as such have a unique content-addressable associative memory where each of the memory locations has dual addresses. For illustration, the proposed algorithm has been used to design a serial-output and a parallel-output 2-bit gray adder.

An enhanced bidirectional associative memory

Abstract: An enhancement of the coding and decoding strategies of a discrete bidirectional associative memory (BAM) reported by B. Kosko (1987, 1988) is presented. Use of the concept, identified as multiple training, can be guaranteed to achieve recall of a single trained pair under suitable initial conditions of data. In representative computer simulations the concept has been shown to represent an improvement in some instances over the original Kosko strategy for recall of multiple pairs. The use of the concept is illustrated, and theorems underlying the result are presented. >

Image processing device

Abstract: PURPOSE:To efficiently reverse an original document upside down when a user sets the original upside down accidentally on an image reader by inverting the sequence of a bit storing of one line before performing compressive encoding, and storing compressively encoded image data at every line in the reverse sequence of readout sequence CONSTITUTION:When an indication for performing an upside-down processing is issued to an encoding line and reference line inversion processing part 13, the encoding line and reference lie inversion processing part 13 sets original image line data l1 in a line memory 3 as an encoding line, and sets original line data l2 in a line memory 4 as a reference line An encoder 5 inputs data of two lines by the instruction of a control part 9 via main scan inversion processing parts 11 and 12, and executes two-dimensional encoding The code compression data D1 of the line data l1 generated by the encoder 5 is stored in the termination of a code compression data buffer memory 6 by a line sequence inversion processing part 14 Since it is possible to perform one- dimensional encoding and the two-dimensional encoding from a line inputted first by reading scan by such upside-down processing, the capacity of a memory can be reduced

Implicit and explicit memories of odors

Abstract: The present experiments explore whether there may be some forms of implicit memory for odors. In the first experiment, the elaborateness of olfactory encoding was varied at presentation. For (explicit) recognition memory testing we found positive effects of labeling responses to odors at encoding. Implicit memory measures (temporal and preference judgments) did not reveal reliable effects of prior odor presentation, however. The second experiment corroborated effects of levels of processing on r recognition memory. Again, perceptual or affective judgments remained insensitive for prior odor exposures. Implicit memory could only be detected with verbal measures at the testing stage (labeling accuracy or latency). These results are consistent with the proposal that odor information is represented at different levels of processing that are even with implicit memory measures only partly accessible.

Circuit of encoding and formatting data for rotating head recorders

Abstract: A circuit for encoding and formatting data, the circuit having an input suitable for receiving a data stream in the form of bytes, the circuit comprising firstly a working memory in which the data is stored in the form of a matrix, and from which the data is taken sequentially, with the data being stored and taken by means of respective different sequences (either by being stored in lines and then taken in columns, or else by being stored in columns and then taken in lines), and secondly first and second encoder means each suitable for calculating, respectively for each line and for each column, a given number of redundancy bytes which are respectively horizontal redundancy bytes and vertical redundancy bytes (with one of said encoding means being connected to the outlet of said memory), wherein the first and second encoding means are connected in parallel to the output from the memory, with said second encoding means (for performing redundancy calculations corresponding to the sequence opposite to that by which the data is taken from the memory) being suitable for storing intermediate results in the calculations performed thereby.

Inter-frame encoding device

Abstract: PURPOSE:To suppress the influence of a transmission line error or an omission to the minimum, to execute the random reproduction at the time when this device is applied to storage device, and also, the obtain the encoding device having a high performance by updating periodically in-frame encoding data, and executing inter-frame encoding between this data and the succeeding frame. CONSTITUTION:Between an encoder 5 and a motion compensating part 14, a switch 20 is provided, the switch 20 is selected to the '0' side at an arbitrary interval at every prescribed frame or in one frame, and in-frame encoding data is sent out to the encoder. Subsequently, the in-frame encoding data 6 obtained by the encoder 5 is brought to variable length encoding, based on its generation frequency together with a motion vector 7 and multiplexed to a series data train by a video multiplexing part 9, and sent out to a cluster constituting part 22. Thereafter, an in-frame decoding processing is executed by a local decoder 8, and an encoding noise is eliminated by a filter 12 and sent out to a switch 21. In this state, the switch 21 is turned on only when the decoder 8 provides an output, and updating of encoded data in an in-loop FM 13 is executed.

Circuit for the estimation of movement in a detected picture

Abstract: In a circuit arrangement for the estimation of movement in a detected picture, video signals arising from the detection being written into at least one picture memory and then subjected to a block-by-block data-reducing encoding and in each case, for the purpose of estimating movement, in each case one block of a picture is compared with comparison blocks in a search field of a preceding picture, a first intermediate memory is provided for the block to be compared in each case and at least a second intermediate memory is provided for the corresponding search field of the preceding picture. The second intermediate memory consists of two read-write memories, in each case data from the picture memory being written into the one and data being read out from the respective other one.

Multiplane image mixing in a display window environment

Abstract: A graphics display system is provided with the ability to use multiple memory buffers to produce images with a wide range of colours through bit plane encoding or to present indeplendent applications displays or combine display images through the use of lateral bit encoding. When operated in the lateral bit encoded state, application programs can be associated with independent memory buffers or an application can use the separate buffers to create a display with animation or apparent movement. Each memory buffer can be independently associated with the display device or the images contained in the memory buffers can be mixed through the use of hardware of software image mixing to create a composite display. The combined image is used to directly control the display device and does not require the creation of an intermediate frame buffer image. This display system provides the capability for animation or image movement through the designation of one or more planes to contain the objects and the designation of display priority among the memory buffers. The image mixer combines the images according to the established display priority so that portions of the highest priority image are always displayed. the images in the memory buffers may, alternatively, be linked to create a single large image which can be scrolled across the display monitor. A method for image mixing and for displaying objects with apparent motion is provided through the use of the multiple memory buffers and display priority assignment.

The Antiphon: A Device for Reliable Memory from Unrealiable Components. I

Abstract: A memory device (the antiphon ) is proposed which reverberates an excitation pattern (corresponding to a state of an M -bit memory) between two sets of nodes (α- and β-nodes). The β-nodes may behave unreliably (stochastically). It is shown that reduced versions of the antiphon produce two memory mechanisms now classical: the repeated cycling of a message through the steps of encoding, noisy transmission and decoding, and the Hopfield net. However, the antiphon presents several differences of detail from the Hopfield net; it is also more physically explicit and more easily analysed. One mode of operation of the antiphon is analysed here; this analysis leads to an optimization of the statistics of the network and a lower bound for the memory capacity of the antiphon.

Implicit memory in spelling from word images.

Abstract: A series of experiments is reported concerning implicit memory in imaginal processing. In the standard condition, subjects had to encode word images before spelling a word. The spelling task was repeated in the test phase with the same words and with additional control words. Spelling times were registered after the image encoding. Implicit memory has been detected if repeated words can be spelled faster than control words. Experiment 1 showed that levels of processing manipulations (such as the additional generation of meaning images at encoding or variations in word concreteness) favor explicit memory, but do not show up in implicit memory. Experiment 2 demonstrated that implicit memory disappears if spelling at encoding took place on visually present words. Experiment 3 investigated whether the focusing of specific letter positions within the image may contribute to the effect, but this was not found. According to a processing view that underlies our task analysis, implicit memory depends on transfer-appropriate processing and is attributed to processes of image encoding or generation and image reconstruction or regeneration.

Optical implementation of A 2-D neural net with a third-order interconnection for image quadratic associative memory based on a compact-encoding method

Abstract: The optical implementations of a 2-D neural net with a third-order interconnection for image quadratic associative memory are reported. Two types of incoherent optical systems, based on multiple imaging and correlation, are used. In particular, a compact-encoding method, which can greatly reduce the size of the interconnection tensor, is also proposed. The experimental results are given.

Image encoding system

Abstract: PURPOSE:To obtain the reproduced image of a satisfactory picture quality with small encoding quantity by dividing an inputted image into a block composed of plural picture elements, after that, selecting one encoder out of the plural encoders based on a parameter to express the signal distribution in the block for every block and executing an encoding. CONSTITUTION:A signal distribution computing part 3 in the block computes the parameter to express a block signal distribution in a buffer memory 2 and the parameter is sent to an encoder selecting part 4. The encoder selecting part 4 selects one encoder out of the plural encoders, whose quantizing characteristics are different, prepared in advance to an encoder memory 5, and then sets it to a variable sampling density encoding/decoding part 6. In the variable sampling density encoding/decoding part 6, an inputted image signal and a forecasted image signal are subtracted by a subtracter 12, a forecasted error signal is obtained and said forecasted error signal is sent to a variable sampling density encoding circuit 10 and a compressing code is outputted. Said compressing code goes to be a reproduced image signal by a variable sampling density decoding circuit 11 and simultaneously is sent to a forecasting circuit 9.

Using back-propagation networks to assess several image representation schemes for object recognition

Abstract: Summary form only given, as follows. Two chapters of research are presented. The first constitutes a demonstration that backpropagation networks can be used as a content addressable memory for visual objects represented within digitized real-world images. For networks encoding two or three classes of traffic signs, classification generalization is demonstrated for objects at new positions on the image frame and also for new instances of a trained class of object. The new instance may even be a somewhat degraded representation. Given this optimistic introduction, the work evolves into a second, more comparative chapter. In this further probe, packpropagation networks are used as content addressable memories with which to determine the relative value of several different visual object representation schemes. These representation schemes are tested along multiple parameters to deduce the efficacy of the scheme itself, and the influence of network parameter changes on the learning and categorization of objects. >

Voice transmission equipment

Abstract: PURPOSE:To improve transmission efficiency when the using frequency of a line is high and to improve sound quality when it is low by using traffic data from an exchange on a transmission side and selecting encoding parts and decoding parts in accordance with a prescribed reference. CONSTITUTION:An encoding part 51 is composed of encoding parts 511-51n where an encoding speed per channel differs, and a decoding part 62 by the decoding parts 621-62n corresponding to the encoding parts. Encoding/decoding part selection parts 53 and 63 are provided in the transmission part/reception part of the equipment. The selection parts 53 and 63 select the encoding parts and corresponding decoding parts in accordance with the prescribed reference by using traffic data from the exchange on the transmission side. Thus, the encoding/decoding parts where the encoding speed is slow are selected and transmission efficiency is improved when the using frequency of the line is high. When the using frequency is low, the encoding/decoding parts where the encoding speed is fast are selected and sound quality is improved.

Digital processing system for video and television signal generation

Abstract: A digital processing system that generates signals for multiple digital video systems and formats including the specific case of PAL D2. The output is either from direct memory or component memory; encoded to composite at line rate, then and scan rate converted to subcarrier rates. The scan rate conversion process includes multiplication process of the signal by coefficients that are changeable with the signal so that conversion error is minimized. A single line of direct memory can be switched in for direct comparison of the outputs to allow detection of scan and encoding errors.

Method and apparatus for interleaved encoding

Abstract: An apparatus and method for interleaved encoding data symbols into code words using a Reed-Solomon error correction code. The invention finds particular utility for the recovery of computer data stored on mass storage media, such as disks.

Efficiency comparison of associative memories using BSC model

Abstract: The efficiency of distributed associative memories in which information is stored redundantly at the expense of memory efficiency, is addressed. The memory use efficiency is evaluated by calculating the mutual information transmitted through a binary symmetric channel (BSC) model, simulating the memorization/recollection process with an encoding/decoding process. Several modifications of the standard matrix-type associative memory that provide higher efficiency of memory usage are proposed. >

Predictive encoder for still pictures

Abstract: A predictive encoder for still pictures comprises a quantizer (Q) whose characteristic curve is variable. The encoded picture data are read and transmitted at a constant bit rate from an output memory (PMUX). A subtractor (SR) and a picture memory (BS) are used for obtaining prediction values. In order to be able to transmit the still pictures from a video camera without any noticeable loss of quality, a first frame memory (V1) is provided in which the data of all pixels of the still picture to be transmitted are stored. A selection circuit (AS) adjusts the characteristic curve after the quantizer (Q) performs the quantizing operation. If the data stored in the first frame memory (V1) are transmitted several times, the selection circuit (AS) selects, for each encoding and transmission operation, a characteristic curve corresponding to a more refined quantizing operation than the characteristic curve which was used for the previous encoding and transmission operation.