Danish Fungi 2020 -- Not Just Another Image Recognition Dataset
This work provides a novel dataset for fine-grained image recognition in mycology, addressing challenges like long-tailed distributions and metadata integration, though it is incremental in benchmarking existing models.
The authors introduced the Danish Fungi 2020 dataset, a fine-grained benchmark with unique features like taxonomy-accurate labels and rich metadata, and found that Vision Transformers outperformed CNNs with 80.45% accuracy and a 0.743 macro F1 score, reducing CNN error by 9-12%, while metadata integration improved accuracy by over 2.95 percentage points, cutting error by 15%.
We introduce a novel fine-grained dataset and benchmark, the Danish Fungi 2020 (DF20). The dataset, constructed from observations submitted to the Atlas of Danish Fungi, is unique in its taxonomy-accurate class labels, small number of errors, highly unbalanced long-tailed class distribution, rich observation metadata, and well-defined class hierarchy. DF20 has zero overlap with ImageNet, allowing unbiased comparison of models fine-tuned from publicly available ImageNet checkpoints. The proposed evaluation protocol enables testing the ability to improve classification using metadata -- e.g. precise geographic location, habitat, and substrate, facilitates classifier calibration testing, and finally allows to study the impact of the device settings on the classification performance. Experiments using Convolutional Neural Networks (CNN) and the recent Vision Transformers (ViT) show that DF20 presents a challenging task. Interestingly, ViT achieves results superior to CNN baselines with 80.45% accuracy and 0.743 macro F1 score, reducing the CNN error by 9% and 12% respectively. A simple procedure for including metadata into the decision process improves the classification accuracy by more than 2.95 percentage points, reducing the error rate by 15%. The source code for all methods and experiments is available at https://sites.google.com/view/danish-fungi-dataset.