Computational platforms for high-performance scientific applications are becoming more heterogenous, including hardware accelerators such as multiple GPUs. Applications in a wide variety of scientific fields require an efficient and careful management of the computational resources of this type of hardware to obtain the best possible performance. However, there are currently different GPU vendors, architectures and families that can be found in heterogeneous clusters or machines. Programming with the vendor provided languages or frameworks, and optimizing for specific devices, may become cumbersome and compromise portability to other systems. To overcome this problem, several proposals for high-level heterogeneous programming have appeared, trying to reduce the development effort and increase functional and performance portability, specifically when using GPU hardware accelerators. This paper evaluates the SYCL programming model, using the Open SYCL compiler, from two different perspectives: The performance it offers when dealing with single or multiple GPU devices from the same or different vendors, and the development effort required to implement the code. We use as case of study the Finite Time Lyapunov Exponent calculation over two real-world scenarios and compare the performance and the development effort of its Open SYCL-based version against the equivalent versions that use CUDA or HIP. Based on the experimental results, we observe that the use of SYCL does not lead to a remarkable overhead in terms of the GPU kernels execution time. In general terms, the Open SYCL development effort for the host code is lower than that observed with CUDA or HIP. Moreover, the SYCL version can take advantage of both CUDA and AMD GPU devices simultaneously much easier than directly using the vendor-specific programming solutions.
相關內容
Noise suppression and echo cancellation are critical in speech enhancement and essential for smart devices and real-time communication. Deployed in voice processing front-ends and edge devices, these algorithms must ensure efficient real-time inference with low computational demands. Traditional edge-based noise suppression often uses MSE-based amplitude spectrum mask training, but this approach has limitations. We introduce a novel projection loss function, diverging from MSE, to enhance noise suppression. This method uses projection techniques to isolate key audio components from noise, significantly improving model performance. For echo cancellation, the function enables direct predictions on LAEC pre-processed outputs, substantially enhancing performance. Our noise suppression model achieves near state-of-the-art results with only 3.1M parameters and 0.4GFlops/s computational load. Moreover, our echo cancellation model outperforms replicated industry-leading models, introducing a new perspective in speech enhancement.
Sequential neural posterior estimation (SNPE) techniques have been recently proposed for dealing with simulation-based models with intractable likelihoods. Unlike approximate Bayesian computation, SNPE techniques learn the posterior from sequential simulation using neural network-based conditional density estimators by minimizing a specific loss function. The SNPE method proposed by Lueckmann et al. (2017) used a calibration kernel to boost the sample weights around the observed data, resulting in a concentrated loss function. However, the use of calibration kernels may increase the variances of both the empirical loss and its gradient, making the training inefficient. To improve the stability of SNPE, this paper proposes to use an adaptive calibration kernel and several variance reduction techniques. The proposed method greatly speeds up the process of training, and provides a better approximation of the posterior than the original SNPE method and some existing competitors as confirmed by numerical experiments.
Advancements in artificial intelligence (AI) over the last decade demonstrate that machines can exhibit communicative behavior and influence how humans think, feel, and behave. In fact, the recent development of ChatGPT has shown that large language models (LLMs) can be leveraged to generate high-quality communication content at scale and across domains, suggesting that they will be increasingly used in practice. However, many questions remain about how knowing the source of the messages influences recipients' evaluation of and preference for AI-generated messages compared to human-generated messages. This paper investigated this topic in the context of vaping prevention messaging. In Study 1, which was pre-registered, we examined the influence of source disclosure on people's evaluation of AI-generated health prevention messages compared to human-generated messages. We found that source disclosure (i.e., labeling the source of a message as AI vs. human) significantly impacted the evaluation of the messages but did not significantly alter message rankings. In a follow-up study (Study 2), we examined how the influence of source disclosure may vary by the participants' negative attitudes towards AI. We found a significant moderating effect of negative attitudes towards AI on message evaluation, but not for message selection. However, for those with moderate levels of negative attitudes towards AI, source disclosure decreased the preference for AI-generated messages. Overall, the results of this series of studies showed a slight bias against AI-generated messages once the source was disclosed, adding to the emerging area of study that lies at the intersection of AI and communication.
Classic cycle-joining techniques have found widespread application in creating universal cycles for a diverse range of combinatorial objects, such as shorthand permutations, weak orders, orientable sequences, and various subsets of $k$-ary strings, including de Bruijn sequences. In the most favorable scenarios, these algorithms operate with a space complexity of $O(n)$ and require $O(n)$ time to generate each symbol in the sequences. In contrast, concatenation-based methods have been developed for a limited selection of universal cycles. In each of these instances, the universal cycles can be generated far more efficiently, with an amortized time complexity of $O(1)$ per symbol, while still using $O(n)$ space. This paper introduces $\mathit{concatenation~trees}$, which serve as the fundamental structures needed to bridge the gap between cycle-joining constructions based on the pure cycle register and corresponding concatenation-based approaches. They immediately demystify the relationship between the classic Lyndon word concatenation construction of de Bruijn sequences and a corresponding cycle-joining based construction. To underscore their significance, concatenation trees are applied to construct universal cycles for shorthand permutations and weak orders in $O(1)$-amortized time per symbol. Moreover, we provide insights as to how similar results can be obtained for other universal cycles including cut-down de Bruijn sequences and orientable sequences.
Nonlinear systems arising from time integrators like Backward Euler can sometimes be reformulated as optimization problems, known as incremental potentials. We show through a comprehensive experimental analysis that the widely used Projected Newton method, which relies on unconditional semidefinite projection of Hessian contributions, typically exhibits a reduced convergence rate compared to classical Newton's method. We demonstrate how factors like resolution, element order, projection method, material model and boundary handling impact convergence of Projected Newton and Newton. Drawing on these findings, we propose the hybrid method Project-on-Demand Newton, which projects only conditionally, and show that it enjoys both the robustness of Projected Newton and convergence rate of Newton. We additionally introduce Kinetic Newton, a regularization-based method that takes advantage of the structure of incremental potentials and avoids projection altogether. We compare the four solvers on hyperelasticity and contact problems. We also present a nuanced discussion of convergence criteria, and propose a new acceleration-based criterion that avoids problems associated with existing residual norm criteria and is easier to interpret. We finally address a fundamental limitation of the Armijo backtracking line search that occasionally blocks convergence, especially for stiff problems. We propose a novel parameter-free, robust line search technique to eliminate this issue.
Interacting systems are ubiquitous in nature and engineering, ranging from particle dynamics in physics to functionally connected brain regions. These interacting systems can be modeled by graphs where edges correspond to the interactions between interactive entities. Revealing interaction laws is of fundamental importance but also particularly challenging due to underlying configurational complexities. The associated challenges become exacerbated for heterogeneous systems that are prevalent in reality, where multiple interaction types coexist simultaneously and relational inference is required. Here, we propose a novel probabilistic method for relational inference, which possesses two distinctive characteristics compared to existing methods. First, it infers the interaction types of different edges collectively, and second, it allows handling systems with variable topological structure over time. We evaluate the proposed methodology across several benchmark datasets and demonstrate that it outperforms existing methods in accurately inferring interaction types. We further show that when combined with known constraints, it allows us, for example, to discover physics-consistent interaction laws of particle systems. Overall the proposed model is data-efficient and generalizable to large systems when trained on smaller ones. The developed methodology constitutes a key element for understanding interacting systems and may find application in graph structure learning.
Despite much attention, the comparison of reduced-dimension representations of high-dimensional data remains a challenging problem in multiple fields, especially when representations remain high-dimensional compared to sample size. We offer a framework for evaluating the topological similarity of high-dimensional representations of very high-dimensional data, a regime where topological structure is more likely captured in the distribution of topological "noise" than a few prominent generators. Treating each representational map as a metric embedding, we compute the Vietoris-Rips persistence of its image. We then use the topological bootstrap to analyze the re-sampling stability of each representation, assigning a "prevalence score" for each nontrivial basis element of its persistence module. Finally, we compare the persistent homology of representations using a prevalence-weighted variant of the Wasserstein distance. Notably, our method is able to compare representations derived from different samples of the same distribution and, in particular, is not restricted to comparisons of graphs on the same vertex set. In addition, representations need not lie in the same metric space. We apply this analysis to a cross-sectional sample of representations of functional neuroimaging data in a large cohort and hierarchically cluster under the prevalence-weighted Wasserstein. We find that the ambient dimension of a representation is a stronger predictor of the number and stability of topological features than its decomposition rank. Our findings suggest that important topological information lies in repeatable, low-persistence homology generators, whose distributions capture important and interpretable differences between high-dimensional data representations.
We propose and compare methods for the analysis of extreme events in complex systems governed by PDEs that involve random parameters, in situations where we are interested in quantifying the probability that a scalar function of the system's solution is above a threshold. If the threshold is large, this probability is small and its accurate estimation is challenging. To tackle this difficulty, we blend theoretical results from large deviation theory (LDT) with numerical tools from PDE-constrained optimization. Our methods first compute parameters that minimize the LDT-rate function over the set of parameters leading to extreme events, using adjoint methods to compute the gradient of this rate function. The minimizers give information about the mechanism of the extreme events as well as estimates of their probability. We then propose a series of methods to refine these estimates, either via importance sampling or geometric approximation of the extreme event sets. Results are formulated for general parameter distributions and detailed expressions are provided when Gaussian distributions. We give theoretical and numerical arguments showing that the performance of our methods is insensitive to the extremeness of the events we are interested in. We illustrate the application of our approach to quantify the probability of extreme tsunami events on shore. Tsunamis are typically caused by a sudden, unpredictable change of the ocean floor elevation during an earthquake. We model this change as a random process, which takes into account the underlying physics. We use the one-dimensional shallow water equation to model tsunamis numerically. In the context of this example, we present a comparison of our methods for extreme event probability estimation, and find which type of ocean floor elevation change leads to the largest tsunamis on shore.
Hashing has been widely used in approximate nearest search for large-scale database retrieval for its computation and storage efficiency. Deep hashing, which devises convolutional neural network architecture to exploit and extract the semantic information or feature of images, has received increasing attention recently. In this survey, several deep supervised hashing methods for image retrieval are evaluated and I conclude three main different directions for deep supervised hashing methods. Several comments are made at the end. Moreover, to break through the bottleneck of the existing hashing methods, I propose a Shadow Recurrent Hashing(SRH) method as a try. Specifically, I devise a CNN architecture to extract the semantic features of images and design a loss function to encourage similar images projected close. To this end, I propose a concept: shadow of the CNN output. During optimization process, the CNN output and its shadow are guiding each other so as to achieve the optimal solution as much as possible. Several experiments on dataset CIFAR-10 show the satisfying performance of SRH.
Deep learning constitutes a recent, modern technique for image processing and data analysis, with promising results and large potential. As deep learning has been successfully applied in various domains, it has recently entered also the domain of agriculture. In this paper, we perform a survey of 40 research efforts that employ deep learning techniques, applied to various agricultural and food production challenges. We examine the particular agricultural problems under study, the specific models and frameworks employed, the sources, nature and pre-processing of data used, and the overall performance achieved according to the metrics used at each work under study. Moreover, we study comparisons of deep learning with other existing popular techniques, in respect to differences in classification or regression performance. Our findings indicate that deep learning provides high accuracy, outperforming existing commonly used image processing techniques.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191