Ensemble of ConvNeXt V2 and MaxViT for Long-Tailed CXR Classification with View-Based Aggregation

Yosuke Yamagishi, Shouhei Hanaoka
10/14/2024
cs.CV

Official code: yamagishi0824/cxrlt24-multiview-pp (2 stars, MIT, last commit 2025-05-01).

Abstract

In this work, we present our solution for the MICCAI 2024 CXR-LT challenge, achieving 4th place in Subtask 2 and 5th in Subtask 1. We leveraged an ensemble of ConvNeXt V2 and MaxViT models, pretrained on an external chest X-ray dataset, to address the long-tailed distribution of chest findings. The proposed method combines state-of-the-art image classification techniques, asymmetric loss for handling class imbalance, and view-based prediction aggregation to enhance classification performance. Through experiments, we demonstrate the advantages of our approach in improving both detection accuracy and the handling of the long-tailed distribution in CXR findings. The code is available at https://github.com/yamagishi0824/cxrlt24-multiview-pp.

View on arXivView PDF

Code Implementations(1)

yamagishi0824/cxrlt24-multiview-ppOfficial100%
MIT
20PythonOct 14, 20241 years agoMIT

Similar Papers

Loss Design and Architecture Selection for Long-Tailed Multi-Label Chest X-Ray Classification

85% match2026

Bag of Tricks for Long-Tailed Multi-Label Classification on Chest X-Rays

85% match2023

Towards long-tailed, multi-label disease classification from chest X-ray: Overview of the CXR-LT challenge

83% match2023

LTCXNet: Advancing Chest X-Ray Analysis with Solutions for Long-Tailed Multi-Label Classification and Fairness Challenges

82% match2024

TRCGL-Net: A Long-Tailed Multi-Label Chest X-Ray Classification Framework with Generative Data Augmentation and Label Co-Occurrence Modeling

82% match2026

Revisiting Long-tailed Image Classification: Survey and Benchmarks with New Evaluation Metrics

81% match2023

Cite this paper

@article{yamagishi2024ensemble,
  title  = {Ensemble of ConvNeXt V2 and MaxViT for Long-Tailed CXR Classification with View-Based Aggregation},
  author = {Yosuke Yamagishi and Shouhei Hanaoka},
  year   = {2024},
  eprint = {2410.10710},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2410.10710v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2410.10710.svg)](https://codeofpaper.com/paper/2410.10710)

Discussion