metadata
license: apache-2.0
datasets:
- Bingsu/Human_Action_Recognition
library_name: transformers
language:
- en
base_model:
- google/siglip2-base-patch16-224
pipeline_tag: image-classification
tags:
- Human-Action-Recognition
Classification Report:
precision recall f1-score support
calling 0.8525 0.7571 0.8020 840
clapping 0.8679 0.7119 0.7822 840
cycling 0.9662 0.9857 0.9758 840
dancing 0.8302 0.8381 0.8341 840
drinking 0.9093 0.8714 0.8900 840
eating 0.9377 0.9131 0.9252 840
fighting 0.9034 0.7905 0.8432 840
hugging 0.9065 0.9000 0.9032 840
laughing 0.7854 0.8583 0.8203 840
listening_to_music 0.8494 0.7988 0.8233 840
running 0.8888 0.9321 0.9099 840
sitting 0.5945 0.7226 0.6523 840
sleeping 0.8593 0.8214 0.8399 840
texting 0.8195 0.6702 0.7374 840
using_laptop 0.6610 0.9190 0.7689 840
accuracy 0.8327 12600
macro avg 0.8421 0.8327 0.8339 12600
weighted avg 0.8421 0.8327 0.8339 12600
