I researched many models , large , medium etc but nano model might not be accurate but it will be good for a great user experience as it uses less resources and it is much faster, choosing a higher accurate model is tradeoff between user experience and accuracy. Also large models will make app heavy use more resources so only small amount of people will be able to experience the ai tagging feature , so I've kept yolov8n model.

I have added an unkown label and changed few paramaters to increase bit of accuracy and reduce wrongly classified images, though it does not drastically increase accuracy , its a minor change.

Currently model can categorize image just in these categories:

class_names = ['person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light',
               'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow',
               'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee',
               'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard',
               'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple',
               'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch',
               'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard',
               'cell phone', 'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase',
               'scissors', 'teddy bear', 'hair drier', 'toothbrush', 'unknown']

To increase classes, I tried training model spent hours finding a dataset but I was not able to find a dataset with bounding labels, my laptop supports training of small nano models so if anyone finds a suitable dataset with many classes and bounding labels do tag me, I'll train the model to add more classes to it , I'll create a seperate issue for training the model.

Fixes #236