IEEE CEC NAS Competition

This repository contains the competition entries, including results, source code, and reports, for the IEEE Congress on Evolutionary Computation (CEC) Competitions on Multiobjective Neural Architecture Search (NAS) for the years 2023 and 2024.

Competitions

IEEE CEC 2024 Competition on Multiobjective Neural Architecture Search

• Overview: The 2024 competition zeroes in on real-time semantic segmentation, a pivotal challenge in autonomous driving systems. Participants are tasked with developing Neural Architecture Search (NAS) solutions that not only segment and classify image pixels quickly and accurately but also meet the stringent real-time processing requirements essential for safe autonomous navigation.
• Details: Competition Website

IEEE CEC 2023 Competition on Multiobjective Neural Architecture Search

• Overview: The 2023 competition is centered on image classification, aiming to enhance NAS approaches that balance computational efficiency with performance. This year’s challenge focuses on advancing evolutionary multiobjective optimization (EMO) algorithms to effectively address the complex multiobjective optimization problems intrinsic to NAS, specifically tailored for the demands of high-performance image classification.
• Details: Competition Website

Repository Structure

/
├── 2023/
│   ├── Data of results/
│   ├── Report of entries/
│   └── Source code of entries/
│   └── Results.xlsx
│   └── Report.pdf
│   └── README.md
└── 2024/
    ├── Data of results/
    ├── Report of entries/
    └── Source code of entries/
    └── Results.xlsx
    └── Report.pdf
    └── README.md

Usage

Each competition year's folder contains three subfolders:

• Data of results/: Contains output files of each competition entries.
• Report of entries/: Contains detailed reports describing the methods, algorithms, and outcomes.
• Source code of entries/: Contains all the source code used for the competition entries.
• Results.xlsx: Contains the statistical results and performance metrics from the competition entries.
• Report.pdf: This report introduces the competition, presents final results, and describes the top three methodologies.

Acknowledgments

Thanks to all the participants and organizers of the IEEE CEC competitions on Multiobjective Neural Architecture Search.