Twenty Snack Food Images for Classification

6,745 images across train, validation, and test splits for classifying 20 different snack foods. The images were sourced from the Google Open Images dataset release 2017_11 and accompany a machine learning tutorial book. It includes categories such as apple, banana, cake, candy, and carrot.

Use Cases

• Train a multi-class image classifier to predict one of the 20 snack food categories.
• Benchmark model performance using the provided train/validation/test splits of 4,838, 955, and 952 images respectively.
• Analyze category distribution across splits, as each category has approximately 350 images.

Strengths

• 6,745 total images provide a substantial sample for model training.
• Clear data splits with 4,838 training, 955 validation, and 952 test images.
• Images are sourced from the established Google Open Images dataset.

Limitations

• The dataset is relatively small for deep learning, with only about 350 images per class.
• Data originates from a 2017 release, which may not reflect current visual styles or products.
• No column or feature-level metadata is provided, limiting analysis beyond image classification.

Provenance

Source

Google Open Images dataset, release 2017_11.

Collection Method

Images were extracted and curated to accompany the book 'Machine Learning by Tutorials'.

Freshness

The dataset card was last updated on 2024-09-10, but the underlying image data is from 2017.