A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
Filters
Programmable Architectures and Design Methods for Two-Variable Numeric Function Generators
2010
IPSJ Transactions on System LSI Design Methodology
Preserved Fulltext
This method can reduce the memory size needed for symmetric functions by nearly half with small speed penalty. The proposed architectures allow a systematic design of various two-variable functions. ...
FPGA implementation results show that, for a complicated function, our NFG achieves 57% of memory size and 60% of delay time of a circuit designed based on a one-variable NFG. 118 ...
Interestingly, for this function, the memory size of the NFGs using uniform segmentation increases in the same way as the memory size of a single look-up table. ...
doi:10.2197/ipsjtsldm.3.118
fatcat:3duwd7ywojfmhgf73zxbbrwnya
Hierarchical Segmentation for Hardware Function Evaluation
2009
IEEE Transactions on Very Large Scale Integration (vlsi) Systems
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2011; you can also visit the original URL.
The file type is application/pdf.
This paper presents a method for evaluating functions based on piecewise polynomial approximations (splines) with a hierarchical segmentation scheme targeting hardware implementation. ...
The hierarchical segmentation method is illustrated using a set of functions including ( 2) log 2 cos 1 ( ) ln( ), a high-degree rational function, ln(1 + ), and 1 (1 + ). ...
BACKGROUND Function evaluation methods can be classified into iterative methods and non-iterative methods. ...
doi:10.1109/tvlsi.2008.2003165
fatcat:xudnulkasrfadcbanmc4w2qcii
Persistent Memory Hash Indexes: An Experimental Evaluation
2021
Proceedings of the VLDB Endowment
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
To this end, this paper provides a comprehensive evaluation of persistent hash tables. ...
Our evaluation was conducted using a unified benchmarking framework and representative workloads. ...
ACKNOWLEDGMENTS This research was supported by the National Key Research and Development Program of China (No. 2018YFB1003502), and by the National Science Foundation of China under Grants 61772183 and ...
dblp:journals/pvldb/HuCW0021
fatcat:xifixwwcirc6vpnpr2aak4od2a
Multi-Scale Squeeze U-SegNet with Multi Global Attention for Brain MRI Segmentation
2021
Sensors
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
In this paper, we propose a multi-scale feature extraction with novel attention-based convolutional learning using the U-SegNet architecture to achieve segmentation of brain tissue from a magnetic resonance ...
The proposed method was evaluated on publicly accessible datasets for brain tissue segmentation. ...
Conflicts of Interest: The authors declare no conflict of interest. ...
doi:10.3390/s21103363
pmid:34066042
fatcat:qlmzlwho3zfjnc4y24fnjxgrlu
A case study of a system level approach to exploration of queuing management schemes for input queue packet switches
2003
Eleventh Euromicro Conference on Parallel, Distributed and Network-Based Processing, 2003. Proceedings.
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2007; you can also visit the original URL.
The file type is application/pdf.
As shown in the experimental results, the method enables us to evaluate the impact of different data management algorithms and different system parameters such as DMA size and packet payload size, in terms ...
The effectiveness of this novel method is demonstrated by a case study of exploring data queuing management schemes for packet switches with input queues. ...
In this way, the amount of details that are simulated at the architectural level is reduced, while a high degree of accuracy is still maintained by appropriately selecting a useful set of functional information ...
doi:10.1109/empdp.2003.1183617
dblp:conf/pdp/HeLJ03
fatcat:pkcdyeemjrac3awp6bsgnunqyq
Image Complexity Guided Network Compression for Biomedical Image Segmentation
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Our proposed framework is shown to be effective for generating compressed segmentation networks, retaining up to ≈ 95% of the full-sized network segmentation accuracy, and at the same time, utilizing ≈ ...
Such mapping enables us to predict the final accuracy for different network sizes, based on the computed dataset complexity. ...
The training uses the Adam [11] optimizer ( 1 = 0.9, 2 = 0.999, = 1e−10) with a fixed learning rate of 0.00002 using a cross-entropy based loss function. ...
arXiv:2107.02927v1
fatcat:zpail5d7gfh73liwf6qtlcfley
Optimal Piecewise Linear Function Approximation for GPU-Based Applications
2016
IEEE Transactions on Cybernetics
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2017; you can also visit the original URL.
The file type is application/pdf.
Other Versions
Following this idea, we propose a novel, efficient, and practical technique to evaluate complex and continuous functions using a nearly optimal design of two types of piecewise linear approximations in ...
the case of a large budget of evaluation subintervals. ...
However, their ease of use comes at the price of non-customizable reduced numerical precision [23] .
A. ...
doi:10.1109/tcyb.2015.2482365
pmid:26462251
fatcat:xzcembv2cncofcnz2x7alr73pm
A Squeeze U-SegNet Architecture Based on Residual Convolution for Brain MRI Segmentation
2022
IEEE Access
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit the original URL.
The file type is application/pdf.
In addition, the method provides a design with more efficient architecture, fewer network parameters, and better segmentation accuracy for brain MRI. ...
The experimental results showed superior performance compared to other stateof-the-art methods on brain MRI segmentation with a dice similarity coefficient (DSC) score of 0.96 and Jaccard index (JI) of ...
Quantitative metrics were used to evaluate the efficacy of a proposed architectural design. ...
doi:10.1109/access.2022.3175188
fatcat:2f4jhdtjfrh6vhotp42atomxum
3D Perception with Slanted Stixels on GPU
2021
IEEE Transactions on Parallel and Distributed Systems
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
This work presents a GPU-accelerated software design of the recently proposed model of Slanted Stixels, which represents the geometric and semantic information of a scene in a compact and accurate way. ...
Our approach achieves real-time performance with high accuracy for 2048 × 1024 image sizes and 4 × 4 Stixel resolution on the low-power embedded GPU of an NVIDIA Tegra Xavier. ...
Moure has focused on the GPU acceleration of Stixels, thus, they acknowledge the support of the project TIN2017-84553-C2-1-R of the Ministerio de Economía, Industria y Competitividad. ...
doi:10.1109/tpds.2021.3067836
fatcat:c2mbt44rvvg4vnmhixzr6xunoi
Automatic segmentation of brain MRI using a novel patch-wise U-net deep architecture
2020
PLoS ONE
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
In the proposed brain segmentation method, the non-overlapping patch-wise U-net is used to overcome the drawbacks of conventional U-net with more retention of local information. ...
In our proposed method, the slices from an MRI scan are divided into non-overlapping patches that are fed into the U-net model along with their corresponding patches of ground truth so as to train the ...
From here, the expansive path starts with up-sampling of the feature maps followed by a 2×2 convolution ("up-convolution") that reduces the number of feature channels to half. ...
doi:10.1371/journal.pone.0236493
pmid:32745102
pmcid:PMC7398543
fatcat:vgrohlkkqvgd7pngnnljmbxwfq
Consistent RDMA-Friendly Hashing on Remote Persistent Memory
[article]
2021
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2021; you can also visit the original URL.
The file type is application/pdf.
Coalescing RDMA and Persistent Memory (PM) delivers high end-to-end performance for networked storage systems, which requires rethinking the design of efficient hash structures. ...
Evaluation results demonstrate that compared with state-of-the-art schemes, continuity hashing achieves high throughput (i.e., 1.45X -- 2.43X improvement), low latency (about 1.7X speedup) and the smallest ...
Specifically, the numbered buckets are addressable by a hash function, and two numbered buckets with adjacent bucket numbers (but non-contiguous memory addresses) share a fine-grained contiguous memory ...
arXiv:2107.06836v1
fatcat:lpvj3qsat5cjpbh7nk6h27yuvu
Efficient virtual memory for big memory servers
2013
SIGARCH Computer Architecture News
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2016; you can also visit the original URL.
The file type is application/pdf.
To remove the TLB miss overhead for big-memory workloads, we propose mapping part of a process's linear virtual address space with a direct segment, while page mapping the rest of the virtual address space ...
Direct segments use minimal hardware-base, limit and offset registers per core-to map contiguous virtual memory regions directly to contiguous physical memory. ...
The views expressed herein are not necessarily those of any sponsor. Hill has a significant financial interest in AMD, and Swift has a significant financial interest in Microsoft. ...
doi:10.1145/2508148.2485943
fatcat:vix4kkpe5veefmas7inuv72uay
Efficient virtual memory for big memory servers
2013
Proceedings of the 40th Annual International Symposium on Computer Architecture - ISCA '13
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2016; you can also visit the original URL.
The file type is application/pdf.
To remove the TLB miss overhead for big-memory workloads, we propose mapping part of a process's linear virtual address space with a direct segment, while page mapping the rest of the virtual address space ...
Direct segments use minimal hardware-base, limit and offset registers per core-to map contiguous virtual memory regions directly to contiguous physical memory. ...
The views expressed herein are not necessarily those of any sponsor. Hill has a significant financial interest in AMD, and Swift has a significant financial interest in Microsoft. ...
doi:10.1145/2485922.2485943
dblp:conf/isca/BasuGCHS13
fatcat:2p7dghs7g5axrn7dh2tttcufoe
NuI-Go: Recursive Non-Local Encoder-Decoder Network for Retinal Image Non-Uniform Illumination Removal
[article]
2020
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2020; you can also visit the original URL.
The file type is application/pdf.
We further demonstrate the advantages of the proposed method for improving the accuracy of retinal vessel segmentation. ...
analysis.To address this issue, we propose a non-uniform illumination removal network for retinal image, called NuI-Go, which consists of three Recursive Non-local Encoder-Decoder Residual Blocks (NEDRBs ...
In [34] , a non-uniform illuminated fundus image enhancement method was proposed, which reduces the blurriness of fundus images based on the cataract physical model and enhances the images with an objective ...
arXiv:2008.02984v1
fatcat:yf77vmdebzhtzdwlnnz6lcvws4
Prioritized Experience Replay
[article]
2016
arXiv
pre-print
Preserved Fulltext
A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2019; you can also visit the original URL.
The file type is application/pdf.
Other Versions
DQN with prioritized experience replay achieves a new state-of-the-art, outperforming DQN with uniform replay on 41 out of 49 games. ...
In prior work, experience transitions were uniformly sampled from a replay memory. ...
Furthermore, Hinton (2007) introduced a form of non-uniform sampling based on error, with an importance sampling correction, which led to a 3x speed-up on MNIST digit classification. ...
arXiv:1511.05952v4
fatcat:mcttbjzpsvhhrkcupyt2cksqai
« Previous
Showing results 1 — 15 out of 32,682 results