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
.
Filters
Computing in the Blink of an Eye: Current Possibilities for Edge Computing and Hardware- Agnostic Programming
2020
IEEE Access
The embedded solutions are compared with respect to their cost, complexity, energy consumption and computing speed establishing valuable guidelines for designers of complex systems that need to make use ...
It is thus critical that such interactions run smoothly in real time, and that the automation strategies do not introduce important delays, usually not larger than 100 milliseconds, as the blink of a human ...
Both these works have reported an improvement in performance and In this work, we have designed an accelerating core for neural network inference task. ...
doi:10.1109/access.2020.2977087
fatcat:3dfxdvm7vrefbju2eynjebdngi
MC-CIM: Compute-in-Memory with Monte-Carlo Dropouts for Bayesian Edge Intelligence
[article]
2021
arXiv
pre-print
Deep neural networks (DNN) with deterministic weights cannot express their prediction uncertainties, thereby pose critical risks for applications where the consequences of mispredictions are fatal such ...
Enhancing the computational efficiency of the method, we discuss a novel CIM module that can perform in-memory probabilistic dropout in addition to in-memory weight-input scalar product to support the ...
Acknowledgement: This work was supported by NSF CA-REER Award (2046435) and a gift funding from Intel. ...
arXiv:2111.07125v1
fatcat:n5zbimhq5vbztcxw7dej2fmv4a
Uncertainty Quantification and Resource-Demanding Computer Vision Applications of Deep Learning
[article]
2022
arXiv
pre-print
Bringing deep neural networks (DNNs) into safety critical applications such as automated driving, medical imaging and finance, requires a thorough treatment of the model's uncertainties. ...
Training deep neural networks is already resource demanding and so is also their uncertainty quantification. ...
Acknowledgements The research leading to these results is in part funded by the German Federal Ministry for Economic Affairs and Climate Action within the project "Methoden und Maßnahmen zur Absicherung ...
arXiv:2205.14917v1
fatcat:mnsszj4wonacpbrkf4ls6tee5q
A Deep Learning Approach to on-Node Sensor Data Analytics for Mobile or Wearable Devices
2017
IEEE journal of biomedical and health informatics
This wealth of information requires efficient methods of classification and analysis where deep learning is a promising technique for large-scale data analytics. ...
We also demonstrate that the computation times for the proposed method are consistent with the constraints of real-time onnode processing on smartphones and a wearable sensor platform. ...
The ActiveMiles dataset and code used in this paper are available at http://hamlyn.doc. ic.ac.uk/activemiles/. For data enquiries, please contact ham-lyn@imperial.ac.uk. ...
doi:10.1109/jbhi.2016.2633287
pmid:28026792
fatcat:ahyqq4c6vfflhan3ul7s4i26pq
Metarobotics for Industry and Society: Vision, Technologies, and Opportunities
2024
IEEE Transactions on Industrial Informatics
Potentials for self-determination, self-efficacy, and work-life-flexibility in robotics-related applications in Society 5.0, Industry 4.0, and Industry 5.0 are outlined. ...
For instance, robot programmers will no longer travel worldwide to plan and test robot motions, even collaboratively. ...
Needs and intentions of workers and robots are inferred to enhance process efficiency through anticipation. ...
doi:10.1109/tii.2023.3337380
fatcat:ijvjy6mjpfehbhrdyf5w4s24be
A Systematic Mapping Study of Italian Research on Workflows
2023
Proceedings of the SC '23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis
ACKNOWLEDGMENTS This work has been supported by ICSC -Centro Nazionale di Ricerca in High-Performance Computing, Big Data and Quantum Computing, funded by European Union -NextGenerationEU. ...
If the training of huge neural models is still anchored to HPC facilities, data pre-processing and inference steps are becoming first-class citizens in geographically distributed Big Data pipelines, preferring ...
Big Data management With the advent of Big Data and the rise of Deep Learning, novel algorithms based on neural networks began to co-exist with standard simulation approaches in large-scale scientific ...
doi:10.1145/3624062.3624285
fatcat:6m3z42hglbhahkcy6huucndal4
FlexSA: Flexible Systolic Array Architecture for Efficient Pruned DNN Model Training
[article]
2020
arXiv
pre-print
FlexSA dynamically reconfigures the systolic array structure and offers multiple sub-systolic operating modes, which are designed for energy- and memory bandwidth-efficient processing of tensors with different ...
To make a systolic array efficient for pruning and training, we propose FlexSA, a flexible systolic array architecture. ...
INTRODUCTION Neural network model pruning is a commonly used technique to make deep learning inference fast and memoryefficient [1] - [4] . ...
arXiv:2004.13027v1
fatcat:6q5aiindzbebzbwixeer4nn7ie
Closed-Loop Neural Prostheses with On-Chip Intelligence: A Review and A Low-Latency Machine Learning Model for Brain State Detection
[article]
2021
arXiv
pre-print
To facilitate a fair comparison and guide design choices among various on-chip classifiers, we propose a new energy-area (E-A) efficiency figure of merit that evaluates hardware efficiency and multi-channel ...
Neural prostheses capable of multi-channel neural recording, on-site signal processing, rapid symptom detection, and closed-loop stimulation are critical to enabling such novel treatments. ...
For example, a deep neural network (DNN) classifier was implemented for emotion detection in children with Autism [100] . ...
arXiv:2109.05848v1
fatcat:m6ib42vqpngb5mal4wkkmuydfm
Cheetah: Optimizing and Accelerating Homomorphic Encryption for Private Inference
[article]
2020
arXiv
pre-print
We evaluate several common neural network models (e.g., ResNet50, VGG16, and AlexNet) and show that plaintext-level HE inference for each is feasible with a custom accelerator consuming 30W and 545mm^2 ...
This paper introduces Cheetah, a set of algorithmic and hardware optimizations for HE DNN inference to achieve plaintext DNN inference speeds. ...
For neural networks, we use HE_Mult with decomposition to compute the partial products since weights are presented in plaintext. ...
arXiv:2006.00505v2
fatcat:egwg4lrzdjbdtkr73znlhocn3e
Automated Deep Learning: Neural Architecture Search Is Not the End
[article]
2022
arXiv
pre-print
This endeavor seeks to minimize the need for human involvement and is best known for its achievements in neural architecture search (NAS), a topic that has been the focus of several surveys. ...
Deep learning (DL) has proven to be a highly effective approach for developing models in diverse contexts, including visual perception, speech recognition, and machine translation. ...
"Automating Analytics for Impact" research programme carried out at the Complex Adaptive Systems Lab, University of Technology Sydney, Australia. ...
arXiv:2112.09245v3
fatcat:dujfh7pzmzbrtdyoshkl4kpbsm
Artemis: Articulated Neural Pets with Appearance and Motion synthesis
[article]
2022
arXiv
pre-print
In this paper, we present ARTEMIS, a novel neural modeling and rendering pipeline for generating ARTiculated neural pets with appEarance and Motion synthesIS. ...
The core of our ARTEMIS is a neural-generated (NGI) animal engine, which adopts an efficient octree-based representation for animal animation and fur rendering. ...
., Ltd. for processing the CGI animals models and motion capture data. Besides, we thank Zhenxiao Yu from Shang-haiTech University for producing a supplementary video. ...
arXiv:2202.05628v2
fatcat:v7f6tg64fvfohmtl7riiecn5fu
Medical Image Analysis using Deep Relational Learning
[article]
2023
arXiv
pre-print
In the past ten years, with the help of deep learning, especially the rapid development of deep neural networks, medical image analysis has made remarkable progress. ...
In this thesis, we propose two novel solutions to this problem based on deep relational learning. ...
Thank them for their support to me and the family. I cannot repay the support from my family, and it is also my biggest motivation to continue scientific research.
Acknowledgements iv ...
arXiv:2303.16099v1
fatcat:kh5umd3h5jbjfa55cywdhozfty
Development of wastewater pipe performance index and performance prediction model
2014
International Journal of Sustainable Materials and Structural Systems
This research presents the life cycle of wastewater pipeline identifying the causes of pipe failure in different phases including design, manufacture, construction, operation and maintenance, and repair ...
The performance index evaluates each iii parameter and combines them mathematically through a weighted summation and a fuzzy inference system that reflects the importance of the various factors. ...
The results from W-PIE show pipes are 0, and 1 which are in excellent, and very good respectively. ...
doi:10.1504/ijsmss.2014.062767
fatcat:zl2brykysnbvxchetac5xzhlae
Adaptive Block Floating-Point for Analog Deep Learning Hardware
[article]
2022
arXiv
pre-print
Analog mixed-signal (AMS) devices promise faster, more energy-efficient deep neural network (DNN) inference than their digital counterparts. ...
We evaluate the effectiveness of ABFP on the DNNs in the MLPerf datacenter inference benchmark – realizing less than 1% loss in accuracy compared to FLOAT32. ...
INTRODUCTION The power consumption and carbon footprint of datacenters used to run compute-intensive deep neural networks (DNNs) have grown substantially in the last decade. ...
arXiv:2205.06287v1
fatcat:njlfxn3c5zh2xgh2p5vv3fdrpy
Towards Performing Image Classification and Object Detection with Convolutional Neural Networks in Autonomous Driving Systems: A Survey (December 2021)
2022
IEEE Access
Due to its single forward pass neural network pipeline, the SqueezeDet architecture is effectively fast, small model sized, accurate and energy efficient. ...
[153] proposed a deep adaptive neural network for multi-label image classification, ML-ANet, which enhances the accuracy performance in cross-domain adaptations. ...
His research focuses on both computer vision tasks for autonomous vehicles with deep learning techniques and optimization methods. ...
doi:10.1109/access.2022.3147495
fatcat:i4xtly3gizck3eorcqknbz4xdm
« Previous
Showing results 1 — 15 out of 213 results