What is it?
Henoamly is a multi-modal for anomaly detection in hebrew scriptures. This model employs a multi-modal approach toward anomaly detection, utilizing cutting-edge technologies such as Vision Transformers (ViT) with their attention to diminutive intricacies, as-well as a highly complex LSTM structure for Optical Character Recgonition. Furhtermore, it is cabable of accurately adjusting preprocessing paramterers to achieve OCR-eligible images.
What does it solve?
Hours go into manual anomaly detection in letters and words in hebrew scriptures, approxamtely seven minutes per page. This model is capable of reducing that amount drastically, even on a low-performace core processing unit.
How fast is it?
The following is a plot for its performace on two different processing units. The first being a medium performance CPU, an Intel i7 8th generation, while the second is a microprocessor, low performance unit.
