input validation

[cho-96]

Fix app.py as below

• Ensure an image or video input is provided (assert image is not None, assert video is not None).
• Validate the model ID (assert model_id in ["yolov10n", "yolov10s", "yolov10m", "yolov10b", "yolov10l", "yolov10x"]).
• Confirm the image size is within the allowable range (assert 320 <= image_size <= 1280).
• Verify the confidence threshold is within the permissible range (assert 0.0 <= conf_threshold <= 1.0).

[caiduoduo12138]

for the inference stage, why the image_size cannot larger than 1280? In my trail, if I use the large resolution, the predict boxes will occur some error. If anyone has the same problem?