modify OCR comment

Author: steven-ja

content/posts/machine learning/deep learning/computer vision/florence/index.md

@@ -330,7 +330,8 @@ print(results)
 
 > {'<OCR_WITH_REGION>': {'quad_boxes': [[143.8125, 146.25, 280.9624938964844, 146.25, 280.9624938964844, 172.25, 143.8125, 172.25], [134.0625, 176.25, 281.9375, 176.25, 281.9375, 202.25, 134.0625, 202.25], [172.73748779296875, 206.25, 284.2124938964844, 206.25, 284.2124938964844, 216.25, 172.73748779296875, 216.25], [150.3125, 238.25, 281.9375, 238.25, 281.9375, 247.25, 150.3125, 247.25], [139.58749389648438, 254.25, 284.2124938964844, 254.25, 284.2124938964844, 277.75, 139.58749389648438, 277.75], [133.08749389648438, 283.75, 285.1875, 283.75, 285.1875, 307.75, 133.08749389648438, 307.75], [140.5625, 312.75, 281.9375, 312.75, 281.9375, 320.75, 140.5625, 320.75]], 'labels': ['</s>**QUANTUM', 'MECHANICS', '(Non-relativistic Theory)', 'Course of Theoretical Phyias Volume 3', 'L.D. LANDAU', 'E.M. LIFSHITZ', 'Initiute of Physical Problems, USSR Academy of**']}}
 
-The overall extracted text from the image is very close to the original one. 
+The overall extracted text from the image is very close to the original one. However, since the image resolution is low, the accuracy on the extracted text is quite low.
+
 ```python
 def draw_ocr_bboxes(image, prediction):
     scale = 1