Fungi MultiClass Microscopic is a high-quality microscopic image dataset released in 2025 for image classification and deep learning research. It aims to provide reliable training and evaluation data resources for fields such as medical mycology and agricultural pathology diagnosis.

This dataset contains 6,801 microscopic images of fungi, covering five fungal categories (H1, H2, H3, H5, and H6). The dataset is divided into training, validation, and test sets. The training set contains 5,000 images with balanced class distribution (1,000 images per class); the validation set contains 899 images with imbalanced class distribution; and the test set contains 902 images, also exhibiting imbalance.

Dataset distribution:

The number of categories is expressed in the format of "training set / validation set / test set".

• H1: 1000 / 437 / 437
• H2: 1000 / 232 / 233
• H3: 1000 / 81 / 81
• H5: 1000 / 80 / 80
• H6: 1000 / 69 / 69