Domain Adaptive Pre-trained Model for Mushroom Image Classification

Yifei SHEN*, Zhuo LI, Yu YANG, Jiaxing SHEN

*Corresponding author for this work

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)Researchpeer-review


Mushroom is highly diverse in morphology and colors, and difficult for ordinary people to discriminate between them. The lack of high-quality labeled mushroom datasets is one of the bottlenecks restricting cutting-edge image recognition models to achieve high performance in mushroom image recognition. To address the limitation, in this paper, we will introduce a large mushroom dataset, called Mushroom-23, constructed by us. It collects over 35000 labeled mushroom images belonging to 203 popular mushroom species in Hong Kong. Along with constructing the new mushroom dataset, we also propose the domain adaptive pre-trained (DAPT) model to make the state-of-the-art Vision Transformer (ViT) adaptive to specific mushroom recognition. The DAPT is first pre-trained on ImageNet and Danish Fungi (DF20) datasets, then fine-tuned on the collected Mushroom-23 dataset to gain the capability of different categories of mushrooms. Extensive experimental results show DAPT outperforms all baseline models by a large margin in terms of Accuracy and Macro F1.

Original languageEnglish
Title of host publicationAdvanced Data Mining and Applications :19th International Conference, ADMA 2023, Shenyang, China, August 21–23, 2023, Proceedings, Part IV
EditorsXiaochun YANG, Heru SUHARTANTO, Guoren WANG, Bin WANG, Jing JIANG, Bing LI, Huaijie ZHU, Ningning CUI
Number of pages13
ISBN (Electronic)9783031466748
ISBN (Print)9783031466731
Publication statusPublished - 2023
Event19th International Conference on Advanced Data Mining and Applications, ADMA 2023 - Shenyang, China
Duration: 21 Aug 202323 Aug 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference19th International Conference on Advanced Data Mining and Applications, ADMA 2023

Bibliographical note

Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.


  • Mushroom Recognition
  • Domain Adaption
  • Pre-trained Model
  • Vision Transformer


Dive into the research topics of 'Domain Adaptive Pre-trained Model for Mushroom Image Classification'. Together they form a unique fingerprint.

Cite this