Multimodal Side-Tuning for Document Classification

Benchmark Model Rank Results
document-image-classification-on-tobacco-3482Multimodal Side-Tuning (MobileNetV2)#2Accuracy: 90.50
document-image-classification-on-tobacco-3482Multimodal Side-Tuning (ResNet50)#3Accuracy: 90.30
document-image-classification-on-rvl-cdipMultimodal (ResNet50)#18Accuracy: 92.7%Parameters: 57M
document-image-classification-on-rvl-cdipMultimodal (MobileNetV2)#22Accuracy: 92.2%Parameters: 12M