Fungi MultiClass Microscopic Fungi Images
Description
A comprehensive image classification dataset containing 6,801 high-quality microscopic images of fungi across 5 distinct classes (H1, H2, H3, H5, H6). The dataset is pre-split into training (5,000 images), validation (899 images), and test (902 images) sets, with balanced class distribution in the training set (1,000 images per class). This dataset is optimized for deep learning model development , This dataset provides a curated collection of microscopic fungi images designed for computer vision and deep learning applications. The dataset has been carefully organized and validated for use in classification tasks, transfer learning experiments, and comparative model analysis. Dataset Characteristics Total Images: 6,801 Number of Classes: 5 (H1, H2, H3, H5, H6) Image Format: JPG Image Type: Microscopic fungi images Dataset Split: Training Set: 5,000 images (1,000 per class) Validation Set: 899 images (variable distribution) Test Set: 902 images (variable distribution) Class Distribution Training Set H1: 1,000 images H2: 1,000 images H3: 1,000 images H5: 1,000 images H6: 1,000 images Validation Set H1: 437 images H2: 232 images H3: 81 images H5: 80 images H6: 69 images Test Set H1: 437 images H2: 233 images H3: 81 images H5: 80 images H6: 69 images Use Cases Deep Learning Model Training: Train and compare various CNN architectures for image classification Transfer Learning: Fine-tune pre-trained models for fungi classification Medical Mycology Research: Support research in fungal identification and classification Agricultural Pathology: Develop automated systems for plant disease detection Computer Vision Benchmarking: Evaluate model performance on microscopic image classification Educational Purposes: Teaching image classification and deep learning concepts Data Quality High-resolution microscopic images Consistent image format (JPG) Pre-split dataset (train/validation/test) Balanced training set for optimal model training Real-world variability in validation and test sets File Information dataset_index.csv: Complete index of all images with split, class, image path, and filename All images are stored in organized class-specific subdirectories File naming convention follows pattern: {Class}_{identifier}.jpg
Files
Steps to reproduce
Dataset Characteristics Total Images: 6,801 Number of Classes: 5 (H1, H2, H3, H5, H6) Image Format: JPG Image Type: Microscopic fungi images Dataset Split: Training Set: 5,000 images (1,000 per class) Validation Set: 899 images (variable distribution) Test Set: 902 images (variable distribution) Class Distribution Training Set H1: 1,000 images H2: 1,000 images H3: 1,000 images H5: 1,000 images H6: 1,000 images Validation Set H1: 437 images H2: 232 images H3: 81 images H5: 80 images H6: 69 images Test Set H1: 437 images H2: 233 images H3: 81 images H5: 80 images H6: 69 images