I have tried to run this model from the link:
https://www.kaggle.com/code/alexfordna/garbage-classification-mobilenetv2-92-accuracy/notebook

When I did it on colab with similar dataset (but smaller, 2100 img to 6 classes), it works well. But when I added this code to predict the input image:

from google.colab import files
from PIL import Image


def process_uploaded_image(image_path, target_size=(224, 224)):
    img = Image.open(image_path)
    img = img.resize(target_size)  
    img_array = np.array(img) 

    if img_array.shape[-1] == 4:  
        img_array = img_array[..., :3]

    img_array = img_array / 255.0 
    img_array = np.expand_dims(img_array, axis=0)  
    img_array = mobilenetv2.preprocess_input(img_array) 

    return img_array

uploaded = files.upload()

for fn in uploaded.keys():  
    processed_image = process_uploaded_image(fn, target_size=IMAGE_SIZE)    
    preds = model.predict(processed_image)
    pred_class = np.argmax(preds, axis=1)
  
    plt.imshow(Image.open(fn))  # Display the uploaded image
    plt.title(f'Predicted class: {categories[pred_class[0]]}')
    plt.axis('off')
    plt.show()
    print(f'File {fn} is predicted as: {categories[pred_class[0]]}')

The result is a wrong prediction. For example, the model always predict my input as a “trash” class. It will change to another class when I stop the runtime, but it still in a wrong prediction too.
I also added this code to check the prediction probabilities:

preds = model.predict(processed_image)
pred_probs = preds[0]  # Get the prediction probabilities for the first (and only) batch
print("Prediction probabilities:", pred_probs)
pred_class = np.argmax(pred_probs)
print("Predicted class:", categories[pred_class])

the output:

**1/1** ━━━━━━━━━━━━━━━━━━━━ **0s** 24ms/step Prediction probabilities: [0.31027108 0.12315894 0.47848797 0.00863316 0.07789086 0.00155797]   
Predicted class: metal  

Why is this happening, and how can my model predict the result correctly?