skip to main content

Title: Efficient NUFFT backpropagation for stochastic sampling optimization
Optimization-based k-space sampling pattern design often involves the Jacobian matrix of non-uniform fast Fourier transform (NUFFT) operations. Previous works relying on auto-differentiation can be time-consuming and less accurate. This work proposes an approximation method using the relationship between exact non-uniform DFT (NDFT) and NUFFT, demonstrating improved results for the sampling pattern optimization problem.
Authors:
; ;
Award ID(s):
1838179
Publication Date:
NSF-PAR ID:
10309645
Journal Name:
Proceeedings International Society of Magnetic Resonance in Medicine
Sponsoring Org:
National Science Foundation
More Like this
  1. With the widely deployment of powerful deep neural network (DNN) into smart, but resource limited IoT devices, many prior works have been proposed to compress DNN in a hardware-aware manner to reduce the computing complexity, while maintaining accuracy, such as weight quantization, pruning, convolution decomposition, etc. However, in typical DNN compression methods, a smaller, but fixed, network structure is generated from a relative large background model for resource limited hardware accelerator deployment. However, such optimization lacks the ability to tune its structure on-the-fly to best fit for a dynamic computing hardware resource allocation and workloads. In this paper, we mainlymore »review two of our prior works [1], [2] to address this issue, discussing how to construct a dynamic DNN structure through either uniform or non-uniform channel selection based sub-network sampling. The constructed dynamic DNN could tune its computing path to involve different number of channels, thus providing the ability to trade-off between speed, power and accuracy on-the-fly after model deployment. Correspondingly, an emerging Spin-Orbit Torque Magnetic Random-Access-Memory (SOT-MRAM) based Processing-In-Memory (PIM) accelerator will also be discussed for such dynamic neural network structure.« less
  2. Matrix completion, the problem of completing missing entries in a data matrix with low-dimensional structure (such as rank), has seen many fruitful approaches and analyses. Tensor completion is the tensor analog that attempts to impute missing tensor entries from similar low-rank type assumptions. In this paper, we study the tensor completion problem when the sampling pattern is deterministic and possibly non-uniform. We first propose an efficient weighted Higher Order Singular Value Decomposition (HOSVD) algorithm for the recovery of the underlying low-rank tensor from noisy observations and then derive the error bounds under a properly weighted metric. Additionally, the efficiency andmore »accuracy of our algorithm are both tested using synthetic and real datasets in numerical simulations.« less
  3. In this work, we present a 3D single-particle tracking system that can apply tailored sampling patterns to selectively extract photons that yield the most information for particle localization. We demonstrate that off-center sampling at locations predicted by Fisher information utilizes photons most efficiently. When performing localization in a single dimension, optimized off-center sampling patterns gave doubled precision compared to uniform sampling. A ~20% increase in precision compared to uniform sampling can be achieved when a similar off-center pattern is used in 3D localization. Here, we systematically investigated the photon efficiency of different emission patterns in a diffraction-limited system and achievedmore »higher precision than uniform sampling. The ability to maximize information from the limited number of photons demonstrated here is critical for particle tracking applications in biological samples, where photons may be limited.« less
  4. Many Markov Chain Monte Carlo (MCMC) methods leverage gradient information of the potential function of target distribution to explore sample space efficiently. However, computing gradients can often be computationally expensive for large scale applications, such as those in contemporary machine learning. Stochastic Gradient (SG-)MCMC methods approximate gradients by stochastic ones, commonly via uniformly subsampled data points, and achieve improved computational efficiency, however at the price of introducing sampling error. We propose a non-uniform subsampling scheme to improve the sampling accuracy. The proposed exponentially weighted stochastic gradient (EWSG) is designed so that a non-uniform-SG-MCMC method mimics the statistical behavior ofmore »a batch-gradient-MCMC method, and hence the inaccuracy due to SG approximation is reduced. EWSG differs from classical variance reduction (VR) techniques as it focuses on the entire distribution instead of just the variance; nevertheless, its reduced local variance is also proved. EWSG can also be viewed as an extension of the importance sampling idea, successful for stochastic-gradient-based optimizations, to sampling tasks. In our practical implementation of EWSG, the non-uniform subsampling is performed efficiently via a Metropolis-Hastings chain on the data index, which is coupled to the MCMC algorithm. Numerical experiments are provided, not only to demonstrate EWSG's effectiveness, but also to guide hyperparameter choices, and validate our non-asymptotic global error bound despite of approximations in the implementation. Notably, while statistical accuracy is improved, convergence speed can be comparable to the uniform version, which renders EWSG a practical alternative to VR (but EWSG and VR can be combined too).

    « less
  5. Recently, the promising aspects of compressive sensing have inspired new circuit-level approaches for their efficient realization within the literature. However, most of these recent advances involving novel sampling techniques have been proposed without considering hardware and signal constraints. Additionally, traditional hardware designs for generating non-uniform sampling clock incur large area overhead and power dissipation. Herein, we propose a novel non-uniform clock generator called Adaptive Quantization Rate (AQR) generator using Magnetic Random Access Memory (MRAM)-based stochastic oscillator devices. Our proposed AQR generator provides orders of magnitude reduction in area while offering 6-fold reduced power dissipation, on average, compared to the state-of-the-artmore »non-uniform clock generators.« less