Returns with about 83% accuracy whether is a crime in a CCTV camera image.

See https://www.kaggle.com/code/dima806/crime-cctv-detection-vit for more details.

Classification report:

              precision    recall  f1-score   support

       Crime     0.7971    0.7823    0.7896     46356
      Normal     0.8467    0.8579    0.8522     64952

    accuracy                         0.8264    111308
   macro avg     0.8219    0.8201    0.8209    111308
weighted avg     0.8260    0.8264    0.8262    111308

Downloads last month: 30

Safetensors

Model size

85.8M params

Tensor type

F32

Inference API

Unable to determine this model's library. Check the docs .

Model tree for dima806/crime_cctv_image_detection

Base model

google/vit-base-patch16-224-in21k

Finetuned

(1676)

this model