Search NASA⌕ Search

SEARCH · Search NASA

Results for “quantization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 217 records · Page 12
All sourcesNASA NTRSDOE OSTI
NASA NTRS2010

Optimal Compression of Floating-Point Astronomical Images Without Significant Loss of Information

We describe a compression method for floating-point astronomical images that gives compression ratios of 6 - 10 while still preserving the scientifically important information in the image. The pixel values are first preprocessed by quantizing them into scaled integer intensity levels, which removes some of the uncompressible noise in the image. The integers are then losslessly compressed using the fast and efficient Rice algorithm and stored in a portable FITS format file. Quantizing an image more coarsely gives greater image compression, but it also increases the noise and degrades the precision of the photometric and astrometric measurements in the quantized image. Dithering the pixel values during the quantization process greatly improves the precision of measurements in the more coarsely quantized images. We perform a series of experiments on both synthetic and real astronomical CCD images to quantitatively demonstrate that the magnitudes and positions of stars in the quantized images can be measured with the predicted amount of precision. In order to encourage wider use of these image compression methods, we have made available a pair of general-purpose image compression programs, called fpack and funpack, which can be used to compress any FITS format image.

Pence, William D.↗
NASA NTRS1975

Three level delta modulation

The basic concept of three-level delta modulation (TLDM) is to quantize the error signal of a delta modulator into three levels instead of the usual two levels as in ordinary DM. A TLDM system must be able to transmit three different signals, corresponding to three possible states of the quantizer, whereas the ordinary DM system has to transmit only two signals. The problem of transmitting these signals over a noisy channel is examined, where the input signal is band-limited Gaussian noise. Two parameters that can be adjusted to optimize the performance of the system with respect to quantizing noise are considered: step size and quantizer threshold setting. The effect of additive channel noise on the performance of a TLDM system is assessed in terms of baseband transmission, probability of error, FSK and PSK transmission, transmission bandwidth, and baseband pulse transmission. It is concluded that TLDM can provide a 5 to 6 dB improvement in quantizing signal to noise ratio over ordinary DM for band-limited Gaussian signals. This improvement can be achieved without increasing the channel bandwidth and with an increase in system threshold of less than 2 dB.

Frank, J.↗
NASA NTRS1994

Improved image decompression for reduced transform coding artifacts

The perceived quality of images reconstructed from low bit rate compression is severely degraded by the appearance of transform coding artifacts. This paper proposes a method for producing higher quality reconstructed images based on a stochastic model for the image data. Quantization (scalar or vector) partitions the transform coefficient space and maps all points in a partition cell to a representative reconstruction point, usually taken as the centroid of the cell. The proposed image estimation technique selects the reconstruction point within the quantization partition cell which results in a reconstructed image which best fits a non-Gaussian Markov random field (MRF) image model. This approach results in a convex constrained optimization problem which can be solved iteratively. At each iteration, the gradient projection method is used to update the estimate based on the image model. In the transform domain, the resulting coefficient reconstruction points are projected to the particular quantization partition cells defined by the compressed image. Experimental results will be shown for images compressed using scalar quantization of block DCT and using vector quantization of subband wavelet transform. The proposed image decompression provides a reconstructed image with reduced visibility of transform coding artifacts and superior perceived quality.

Orourke, Thomas P.↗
DOE OSTI2026

Pythia8 Quark and Gluon Jets (float8 e4m3FN)

A float8 (e4m3FN) quantized version of the quark and gluon jet dataset originally published by Komiske, Metodiev, and Thaler (Zenodo record 3164691). Only the 20-file subset without charm and bottom quark jets is included here. All simulation parameters and jet selection criteria are identical to the original: Pythia 8.226, √s = 14 TeV Quarks from WeakBosonAndParton:qg2gmZq, gluons from WeakBosonAndParton:qqbar2gmZg with the Z decaying to neutrinos FastJet 3.3.0, anti-k_t jets with R = 0.4 p_T^jet ∈ [500, 550] GeV, |y^jet| < 1.7 There are 20 files, each in compressed NumPy format (QG_jets_fp8e4m3fn_0.npz through QG_jets_fp8e4m3fn_19.npz). Each file contains two arrays: X: (100000, M, 4) — 50k quark and 50k gluon jets, randomly sorted, padded to max multiplicity M, with particle features (pt, rapidity, azimuthal angle, pdgid) y: (100000,) — jet labels, gluon = 0, quark = 1 Since NumPy has no native fp8 dtype, X is stored as float32, but the values have been quantized through TensorFlow's float8_e4m3fn type and carry only fp8 precision. The quantization procedure is as follows: a global per-channel scale factor is computed from the absolute maximum value across all 20 chunks (with FP8_MAX = 448.0, the maximum representable value of e4m3FN). Each chunk is then scaled into the fp8 dynamic range, round-tripped through tf.experimental.float8_e4m3fn, and scaled back. This global scaling ensures a consistent quantization grid across the full dataset. The y labels are unchanged. Users should be aware that e4m3FN has limited dynamic range and precision. We recommend verifying this format is appropriate for your application; for a less aggressive reduction see the float16 and float32 versions linked below. If you use this dataset, please cite the original Zenodo record and its associated paper: Komiske, Metodiev, Thaler, Energy Flow Networks: Deep Sets for Particle Jets, JHEP 01 (2019) 121, arXiv:1810.05165

DiLullo, Nicholas [Brown University] (ORCID:000000↗
NASA NTRS1975

Research study on stabilization and control. Modern sampled-data control theory: Stability analysis of the low cost large space telescope system

The pointing stability of the low-cost large space telescope (LST) system was investigated. The low-cost LST is characterized by the use of reaction wheels for the generation of control torques. Because of the critical requirement on the pointing accuracy of the LST, the nonlinear frictional characteristics of the bearings of the reaction wheels were studied which can cause limit cycles in a closed-loop system. Another possible source of pointing error in the LST is due to the effect of quantization and sensor noise. Since the LST is a digital system, digital to analog and analog to digital converters and sensors for positional and rate feedbacks were used. Sensor noise and amplitude quantization causes pointing error in the LST and, in addition, quantization is a nonlinear phenomenon which can cause self-sustained oscillations in the closed-loop system. The dynamic modeling of the single-axis LST is described, and several methods of evaluating the attitude error of the digital LST due to quantization and noise inputs are given.

Kuo, B. C.↗
NASA NTRS1978

Fixed-point error analysis of Winograd Fourier transform algorithms

The quantization error introduced by the Winograd Fourier transform algorithm (WFTA) when implemented in fixed-point arithmetic is studied and compared with that of the fast Fourier transform (FFT). The effect of ordering the computational modules and the relative contributions of data quantization error and coefficient quantization error are determined. In addition, the quantization error introduced by the Good-Winograd (GW) algorithm, which uses Good's prime-factor decomposition for the discrete Fourier transform (DFT) together with Winograd's short length DFT algorithms, is studied. Error introduced by the WFTA is, in all cases, worse than that of the FFT. In general, the WFTA requires one or two more bits for data representation to give an error similar to that of the FFT. Error introduced by the GW algorithm is approximately the same as that of the FFT.

Patterson, R. W.↗
NASA NTRS1995

Image data compression having minimum perceptual error

A method for performing image compression that eliminates redundant and invisible image components is described. The image compression uses a Discrete Cosine Transform (DCT) and each DCT coefficient yielded by the transform is quantized by an entry in a quantization matrix which determines the perceived image quality and the bit rate of the image being compressed. The present invention adapts or customizes the quantization matrix to the image being compressed. The quantization matrix comprises visual masking by luminance and contrast techniques and by an error pooling technique all resulting in a minimum perceptual error for any given bit rate, or minimum bit rate for a given perceptual error.

Watson, Andrew B.↗
NASA NTRS1996

DCTune Perceptual Optimization of Compressed Dental X-Rays

In current dental practice, x-rays of completed dental work are often sent to the insurer for verification. It is faster and cheaper to transmit instead digital scans of the x-rays. Further economies result if the images are sent in compressed form. DCTune is a technology for optimizing DCT (digital communication technology) quantization matrices to yield maximum perceptual quality for a given bit-rate, or minimum bit-rate for a given perceptual quality. Perceptual optimization of DCT color quantization matrices. In addition, the technology provides a means of setting the perceptual quality of compressed imagery in a systematic way. The purpose of this research was, with respect to dental x-rays, 1) to verify the advantage of DCTune over standard JPEG (Joint Photographic Experts Group), 2) to verify the quality control feature of DCTune, and 3) to discover regularities in the optimized matrices of a set of images. We optimized matrices for a total of 20 images at two resolutions (150 and 300 dpi) and four bit-rates (0.25, 0.5, 0.75, 1.0 bits/pixel), and examined structural regularities in the resulting matrices. We also conducted psychophysical studies (1) to discover the DCTune quality level at which the images became 'visually lossless,' and (2) to rate the relative quality of DCTune and standard JPEG images at various bitrates. Results include: (1) At both resolutions, DCTune quality is a linear function of bit-rate. (2) DCTune quantization matrices for all images at all bitrates and resolutions are modeled well by an inverse Gaussian, with parameters of amplitude and width. (3) As bit-rate is varied, optimal values of both amplitude and width covary in an approximately linear fashion. (4) Both amplitude and width vary in systematic and orderly fashion with either bit-rate or DCTune quality; simple mathematical functions serve to describe these relationships. (5) In going from 150 to 300 dpi, amplitude parameters are substantially lower and widths larger at corresponding bit-rates or qualities. (6) Visually lossless compression occurs at a DCTune quality value of about 1. (7) At 0.25 bits/pixel, comparative ratings give DCTune a substantial advantage over standard JPEG. As visually lossless bit-rates are approached, this advantage of necessity diminishes. We have concluded that DCTune optimized quantization matrices provide better visual quality than standard JPEG. Meaningful quality levels may be specified by means of the DCTune metric. Optimized matrices are very similar across the class of dental x-rays, suggesting the possibility of a 'class-optimal' matrix. DCTune technology appears to provide some value in the context of compressed dental x-rays.

Watson, Andrew B.↗
NASA NTRS2005

Improved Compression of Wavelet-Transformed Images

A recently developed data-compression method is an adaptive technique for coding quantized wavelet-transformed data, nominally as part of a complete image-data compressor. Unlike some other approaches, this method admits a simple implementation and does not rely on the use of large code tables. A common data compression approach, particularly for images, is to perform a wavelet transform on the input data, and then losslessly compress a quantized version of the wavelet-transformed data. Under this compression approach, it is common for the quantized data to include long sequences, or runs, of zeros. The new coding method uses prefixfree codes for the nonnegative integers as part of an adaptive algorithm for compressing the quantized wavelet-transformed data by run-length coding. In the form of run-length coding used here, the data sequence to be encoded is parsed into strings consisting of some number (possibly 0) of zeros, followed by a nonzero value. The nonzero value and the length of the run of zeros are encoded. For a data stream that contains a sufficiently high frequency of zeros, this method is known to be more effective than using a single variable length code to encode each symbol. The specific prefix-free codes used are from two classes of variable-length codes: a class known as Golomb codes, and a class known as exponential-Golomb codes. The codes within each class are indexed by a single integer parameter. The present method uses exponential-Golomb codes for the lengths of the runs of zeros, and Golomb codes for the nonzero values. The code parameters within each code class are determined adaptively on the fly as compression proceeds, on the basis of statistics from previously encoded values. In particular, a simple adaptive method has been devised to select the parameter identifying the particular exponential-Golomb code to use. The method tracks the average number of bits used to encode recent runlengths, and takes the difference between this average length and the code parameter. When this difference falls outside a fixed range, the code parameter is updated (increased or decreased). The Golomb code parameter is selected based on the average magnitude of recently encoded nonzero samples. The coding method requires no floating- point operations, and more readily adapts to local statistics than other methods. The method can also accommodate arbitrarily large input values and arbitrarily long runs of zeros. In practice, this means that changes in the dynamic range or size of the input data set would not require a change to the compressor. The algorithm has been tested in computational experiments on test images. A comparison with a previously developed algorithm that uses large code tables (generated via Huffman coding on training data) suggests that the data-compression effectiveness of the present algorithm is comparable to the best performance achievable by the previously developed algorithm.

Kiely, Aaron↗
DOE OSTI2025

Edge-driven transition between extended quantum anomalous Hall crystal and fractional Chern insulator in rhombohedral graphene multilayers

Fractional Chern insulators (FCI) with fractionally quantized Hall conductance at fractional fillings and an extended quantum anomalous Hall (EQAH) crystal with an integer quantized Hall conductance over an extended region of doping were recently observed in pentalayer graphene. One particularly puzzling observation is the transition between the EQAH and FCI regimes, driven either by temperature or electrical current. Here, in this study, we propose a scenario to understand these transitions based on the topologically protected gapless edge modes that are present in both the FCI and EQAH phases and should be most relevant at temperature scales below the energy gap. Our consideration is based on the simple assumption that the edge velocity in FCI is smaller than that in EQAHE and thus contributes to a higher entropy. We further argue that domains with opposite fractionally quantized Hall conductance are ubiquitous in the devices due to disorder, which gives rise to a network of edge modes. The velocity of the edge modes between domains is further reduced due to edge reconstruction. The edge velocity can also be reduced by current when the occupation of the edge mode approaches the gap edge. The edge entropy therefore drives the transition from EQAH to FCI either by temperature or current at a nonzero temperature.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

Perturbative quantum evolution of the gravitational state and dressing in general backgrounds

This paper sets up a perturbative treatment of the evolving quantum state of a gravitational system, in a Schrödinger-like picture, working about a general background. This connects gauge symmetry, the constraints, gravitational dressing, and evolution. Starting with a general time slicing, we give a simple derivation of the relation between the constraints, the Hamiltonian, and its well-known boundary term. Among different approaches to quantization with constraints, we focus on a “gauge-invariant canonical quantization,” which is developed perturbatively in the gravitational coupling. The leading-order solution of the constraints (including the Wheeler-DeWitt equation) for perturbations about the background is given in terms of an explicit construction of gravitational dressings built using certain generalized Green’s functions; different such dressings corresponding to adding propagating gravitational waves to a particular solution of the constraints. Dressed operators commute with the constraints, expressing their gauge invariance, and have an algebraic structure differing significantly from the undressed operators of the underlying field theory. These operators can act on the vacuum to create dressed states, and evolution of general such states is then generated by the boundary Hamiltonian, and alternately may be characterized using other relational observables. This provides a concrete approach to studying perturbative time evolution, including the leading gravitational backreaction, of quantum states of black holes with flat or anti–de-Sitter asymptotics, for example on horizon-crossing slices. This description of evolution in turn provides a starting point for investigating possibly important corrections to quantum evolution, that go beyond quantized general relativity. Published by the American Physical Society 2024

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Chaotic roots of the modular multiplication dynamical system in Shor's algorithm

Shor's factoring algorithm, believed to provide an exponential speedup over classical computation, relies on finding the period of an exactly periodic quantum modular multiplication operator. This exact periodicity is the hallmark of an integrable system, which is paradoxical from the viewpoint of quantum chaos, given that the classical limit of the modular multiplication operator is a highly chaotic system that occupies the “maximally random” Bernoulli level of the classical ergodic hierarchy. In this work, we approach this apparent paradox from a quantum dynamical systems viewpoint, and consider whether signatures of ergodicity and chaos may indeed be encoded in such an “integrable” quantization of a chaotic system. We show that Shor's modular multiplication operator, in specific cases, can be written as a superposition of quantized 𝐴-baker's maps exhibiting more typical signatures of quantum chaos and ergodicity. This work suggests that the integrability of Shor's modular multiplication operator may stem from the interference of other “chaotic” quantizations of the same family of maps, and paves the way for deeper studies on the interplay of integrability, ergodicity, and chaos in and via quantum algorithms.

Dynamical systems↗
DOE OSTI2024

Lion Cub: Minimizing Communication Overhead in Distributed Lion

Communication overhead is a key challenge in distributed deep learning, especially on slower Ethernet intercon nects, and given current hardware trends, communication is likely to become a major bottleneck. While gradient compression techniques have been explored for SGD and Adam, the Lion optimizer has the distinct advantage that its update vectors are the output of a sign operation, enabling straightforward quantization. However, simply compressing updates for communication and using techniques like majority voting fails to lead to end-to-end speedups due to inefficient communication algorithms and reduced convergence. We analyze three factors critical to distributed learning with Lion: optimizing communication methods, identifying effective quantization methods, and assessing the necessity of momentum synchronization. Our findings show that quantization techniques adapted to Lion and selective momentum synchronization can significantly reduce communication costs while maintaining convergence. We combine these into Lion Cub, which enables up to 5x speedups in end-to-end training compared to Lion. This highlights Lion’s potential as a communication-efficient solution for distributed training.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1971

Performance of convolution coding concatenated with MFSK modulation in a Gaussian channel

The improvement in db due to concatenation over conventional M-ary coding is studied to reduce the probability of a bit error and to increase the available bit rate for the same system parameters of error rate, transmitter power, and range. The results of calculations for orthogonal modulation with noncoherent detection and Q-level correlator quantization are presented. It is shown that the correlator outputs are quantized to one of the Q levels, and the receiver output is a vector consisting of a list of the M correlator quantum levels. The channel has Q(M) possible outputs and M possible inputs. Optimum output is approached by increasing fine quantization

Choudhury, A. K.↗
NASA NTRS1972

Control system optimization studies. Volume 2: High frequency cutoff filter analysis

The problem of digital implementation of a cutoff filter is approached with consideration to word length, sampling rate, accuracy requirements, computing time and hardware restrictions. Computing time and hardware requirements for four possible programming forms for the linear portions of the filter are determined. Upper bounds for the steady state system output error due to quantization for digital control systems containing a digital network programmed both in the direct form and in the canonical form are derived. This is accomplished by defining a set of error equations in the z domain and then applying the final value theorem to the solution. Quantization error was found to depend upon the digital word length, sampling rate, and system time constants. The error bound developed may be used to estimate the digital word length and sampling rate required to achieve a given system specification. From the quantization error accumulation, computing time and hardware point of view, and the fact that complex poles and zeros must be realized, the canonical form of programming seems preferable.

Fong, M. H.↗
NASA NTRS1976

Digital television system design study

The use of digital techniques for transmission of pictorial data is discussed for multi-frame images (television). Video signals are processed in a manner which includes quantization and coding such that they are separable from the noise introduced into the channel. The performance of digital television systems is determined by the nature of the processing techniques (i.e., whether the video signal itself or, instead, something related to the video signal is quantized and coded) and to the quantization and coding schemes employed.

Huth, G. K.↗
NASA NTRS1976

Evaluation of NASA speech encoder

Techniques developed by NASA for spaceflight instrumentation were used in the design of a quantizer for speech-decoding. Computer simulation of the actions of the quantizer was tested with synthesized and real speech signals. Results were evaluated by a phometician. Topics discussed include the relationship between the number of quantizer levels and the required sampling rate; reconstruction of signals; digital filtering; speech recording, sampling, and storage, and processing results.

Source record↗
NASA NTRS1979

Radiometric resolution for monitoring vegetation: How many bits are needed?

The significance of the various number of radiometric quantizing levels required for satellite monitoring of vegetation resources was evaluated by using in situ collected spectral reflectance data, an atmospheric radiative transfer simulation model, and a satellite sensor simulation model. Reflectance data were converted to radiance data; passed through a model atmosphere to an altitude of 706 km; and subsequently quantized at 16, 32, 64, 128, 256, and 512 digital count levels for Thematic Mapper bands TM3(0.63 - 0.69 microns) and TM4(0.76 - 0.90 microns). The simulated digital count data were regressed against the in situ biological data to quantify the relationship between quantizing levels.

Tucker, C. J.↗