Results 81 to 90 of about 254 (215)
Capacitive, charge‐domain compute‐in‐memory (CIM) stores weights as capacitance,eliminating DC sneak paths and IR‐drop, yielding near‐zero standbypower. In this perspective, we present a device to systems level performance analysis of most promising architectures and predict apathway for upscaling capacitive CIM for sustainable edge computing ...
Kapil Bhardwaj +2 more
wiley +1 more source
Contention Resolution with Heterogeneous Job Sizes [PDF]
We study the problem of contention resolution for different-sized jobs on a simple channel. When a job makes a run attempt, it learns only whether the attempt succeeded or failed. We first analyze binary exponential backoff, and show that it achieves a makespan of V2Θ(√logn) with high probability, where V is the total work of all n contending jobs ...
Michael A. Bender +2 more
openaire +1 more source
Strategic Contention Resolution in Multiple Channels [PDF]
The results of this work are included in the 11th International Symposium on Algorithmic Game Theory (SAGT 2018) and the 16th Workshop on Approximation and Online Algorithms (WAOA 2018)
Christodoulou, George +2 more
openaire +3 more sources
Variational Autoencoder+Deep Deterministic Policy Gradient addresses low‐light failures of infrared depth sensing for indoor robot navigation. Stage 1 pretrains an attention‐enhanced Variational Autoencoder (Convolutional Block Attention Module+Feature Pyramid Network) to map dark depth frames to a well‐lit reconstruction, yielding a 128‐D latent code ...
Uiseok Lee +7 more
wiley +1 more source
The Watershed as a Conceptual Framework for the Study of Environmental and Human Health
The watershed provides a physical basis for establishing linkages between aquatic contaminants, environmental health and human health. Current attempts to establish such linkages are limited by environmental and epidemiological constraints. Environmental
Alan S. Kolok +3 more
doaj +2 more sources
Illustration of text data mining of rare earth mineral thermodynamic parameters with the large language model‐powered LMExt. A dataset is built with mined thermodynamic properties. Subsequently, a machine learning model is trained to predict formation enthalpy from the dataset.
Juejing Liu +6 more
wiley +1 more source
The use of Low Power Wide Area (LPWA) networks in the 920-MHz band, which are compatible with terrestrial network devices, is being explored for a satellite Internet-of-Things (IoT) system.
Hiraku Okada +6 more
doaj +1 more source
Feature from recent image foundation models (DINOv2) are useful for vision tasks (segmentation, object localization) with little or no human input. Once upsampled, they can be used for weakly supervised micrograph segmentation, achieving strong results when compared to classical features (blurs, edge detection) across a range of material systems.
Ronan Docherty +2 more
wiley +1 more source
An analysis of a contention resolution algorithm
A single multiaccess channel is studied with the outcome of a transmission being either 'idle', 'success', or 'collision' (ternary channel). Packets involved in a collision must be retransmitted, and an efficient way to solve a collision is known in the literature as Gallager-Tsybakov-Mikhailov algorithm.
openaire +3 more sources
This review explores the transformative impact of artificial intelligence on multiscale modeling in materials research. It highlights advancements such as machine learning force fields and graph neural networks, which enhance predictive capabilities while reducing computational costs in various applications.
Artem Maevskiy +2 more
wiley +1 more source

