☆ 3.8 Proceedings Paper

A data-stream TinyML compression algorithm for vehicular applications: a case study

PROCEEDINGS OF 2022 IEEE INTERNATIONAL WORKSHOP ON METROLOGY FOR INDUSTRY 4.0 & IOT (IEEE METROIND4.0&IOT) (2022)

A new task offloading algorithm in edge computing

Zhenjiang Zhang et al.

Summary: The Internet of Things is changing the world, and edge computing is an application that can help reduce user latency, enhance user experience, and balance resource utilization. Through deep reinforcement learning algorithms, reasonable task allocation can be achieved to improve system adaptability and stability.

EURASIP JOURNAL ON WIRELESS COMMUNICATIONS AND NETWORKING (2021)

添加到收藏夹

Article Chemistry, Analytical

A Generalization Performance Study Using Deep Learning Networks in Embedded Systems

Joseba Gorospe et al.

Summary: Deep learning techniques are increasingly being used in the scientific community due to the high computational capacity of current systems and the increase in available data. The development of edge computing allows artificial intelligence to be integrated as close as possible to the client, enabling real-time operation without transferring all data to centralized servers.

SENSORS (2021)

添加到收藏夹

Article Chemistry, Analytical

An Evolving TinyML Compression Algorithm for IoT Environments Based on Data Eccentricity

Gabriel Signoretti et al.

Summary: This paper introduces a data compression solution for the Internet of Things based on the perspective of TinyML, called Tiny Anomaly Compressor (TAC). The TAC algorithm, which does not require previously established mathematical models or assumptions about data distribution, showed promising results, achieving a maximum compression rate of 98.33% and surpassing other algorithms in terms of compression error and peak signal-to-noise ratio in all cases.

SENSORS (2021)

添加到收藏夹

Proceedings Paper Computer Science, Information Systems