Spaces:
Runtime error
Runtime error
Commit
·
6104a90
1
Parent(s):
d9ddda8
Use our examples
Browse files
app.py
CHANGED
@@ -11,10 +11,10 @@ model = VisionEncoderDecoderModel.from_pretrained("microsoft/trocr-base-printed"
|
|
11 |
model.config.eos_token_id = 2
|
12 |
|
13 |
# load image examples
|
14 |
-
urls = ['https://layoutlm.blob.core.windows.net/trocr/dataset/SROIE2019Task2Crop/train/X00016469612_1.jpg', 'https://layoutlm.blob.core.windows.net/trocr/dataset/SROIE2019Task2Crop/train/X51005255805_7.jpg', 'https://layoutlm.blob.core.windows.net/trocr/dataset/SROIE2019Task2Crop/train/X51005745214_6.jpg']
|
15 |
-
for idx, url in enumerate(urls):
|
16 |
-
image = Image.open(requests.get(url, stream=True).raw)
|
17 |
-
image.save(f"image_{idx}.png")
|
18 |
|
19 |
def process_image(image):
|
20 |
# prepare image
|
@@ -31,7 +31,9 @@ def process_image(image):
|
|
31 |
title = "Interactive demo: TrOCR"
|
32 |
description = "Demo for Microsoft's TrOCR, an encoder-decoder model consisting of an image Transformer encoder and a text Transformer decoder for state-of-the-art optical character recognition (OCR) on single-text line images. This particular model is fine-tuned on SROIE Task 2, a dataset of annotated printed images. To use it, simply upload a (single-text line) image or use one of the example images below and click 'submit'. Results will show up in a few seconds."
|
33 |
article = "<p style='text-align: center'><a href='https://arxiv.org/abs/2109.10282'>TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models</a> | <a href='https://github.com/microsoft/unilm/tree/master/trocr'>Github Repo</a></p>"
|
34 |
-
examples =[["image_0.png"], ["image_1.png"], ["image_2.png"]]
|
|
|
|
|
35 |
|
36 |
#css = """.output_image, .input_image {height: 600px !important}"""
|
37 |
|
|
|
11 |
model.config.eos_token_id = 2
|
12 |
|
13 |
# load image examples
|
14 |
+
#urls = ['https://layoutlm.blob.core.windows.net/trocr/dataset/SROIE2019Task2Crop/train/X00016469612_1.jpg', 'https://layoutlm.blob.core.windows.net/trocr/dataset/SROIE2019Task2Crop/train/X51005255805_7.jpg', 'https://layoutlm.blob.core.windows.net/trocr/dataset/SROIE2019Task2Crop/train/X51005745214_6.jpg']
|
15 |
+
#for idx, url in enumerate(urls):
|
16 |
+
# image = Image.open(requests.get(url, stream=True).raw)
|
17 |
+
# image.save(f"image_{idx}.png")
|
18 |
|
19 |
def process_image(image):
|
20 |
# prepare image
|
|
|
31 |
title = "Interactive demo: TrOCR"
|
32 |
description = "Demo for Microsoft's TrOCR, an encoder-decoder model consisting of an image Transformer encoder and a text Transformer decoder for state-of-the-art optical character recognition (OCR) on single-text line images. This particular model is fine-tuned on SROIE Task 2, a dataset of annotated printed images. To use it, simply upload a (single-text line) image or use one of the example images below and click 'submit'. Results will show up in a few seconds."
|
33 |
article = "<p style='text-align: center'><a href='https://arxiv.org/abs/2109.10282'>TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models</a> | <a href='https://github.com/microsoft/unilm/tree/master/trocr'>Github Repo</a></p>"
|
34 |
+
#examples =[["image_0.png"], ["image_1.png"], ["image_2.png"]]
|
35 |
+
examples = [["example-line-0.png"], ["example-line-1.png"],["example-line-2.png"]]
|
36 |
+
|
37 |
|
38 |
#css = """.output_image, .input_image {height: 600px !important}"""
|
39 |
|