Spaces:
Sleeping
Sleeping
First commit.
Browse files- SpellingList3.jpg +0 -0
- SpellingTest8.jpg +0 -0
- app.py +81 -0
- requirements.txt +4 -0
SpellingList3.jpg
ADDED
SpellingTest8.jpg
ADDED
app.py
ADDED
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
from paddleocr import PaddleOCR, draw_ocr
|
3 |
+
import asyncio
|
4 |
+
import edge_tts
|
5 |
+
|
6 |
+
def image2Text(image:str, langChoice:str):
|
7 |
+
ocr = PaddleOCR(use_angle_cls=True, lang=langChoice) # need to run only once to download and load model into memory
|
8 |
+
img_path = image
|
9 |
+
result = ocr.ocr(img_path, cls=True)
|
10 |
+
text = ""
|
11 |
+
for idx in range(len(result)):
|
12 |
+
res = result[idx]
|
13 |
+
for line in res:
|
14 |
+
import re
|
15 |
+
# remove pinyin if it's Chinese
|
16 |
+
if langChoice=="ch":
|
17 |
+
#t = re.sub('[a-z0-9.]', '', line[1][0])
|
18 |
+
t = re.sub('[a-z]', '', line[1][0])
|
19 |
+
t = re.sub('[0-9]\.', '', t)
|
20 |
+
t = t.replace(" ", "")
|
21 |
+
t = t.replace("()", "")
|
22 |
+
t = t.replace("()", "")
|
23 |
+
t = t.replace("( )", "")
|
24 |
+
t = t.replace("()", "")
|
25 |
+
if t!="":
|
26 |
+
text +=((t) + "\n")
|
27 |
+
else:
|
28 |
+
t = line[1][0]
|
29 |
+
t = re.sub('Term [0-9] Spelling', '', t)
|
30 |
+
t = re.sub('Page [0-9]', '', t)
|
31 |
+
if t!="":
|
32 |
+
text += (t + "\n")
|
33 |
+
return text
|
34 |
+
|
35 |
+
|
36 |
+
def text2Voice(recognized_text, langChoice:str):
|
37 |
+
async def amain() -> None:
|
38 |
+
"""Main function"""
|
39 |
+
communicate = edge_tts.Communicate(TEXT, VOICE)
|
40 |
+
await communicate.save(OUTPUT_FILE)
|
41 |
+
TEXT = recognized_text
|
42 |
+
if langChoice == "ch":
|
43 |
+
VOICE = "zh-CN-YunxiaNeural"
|
44 |
+
else:
|
45 |
+
VOICE = "en-GB-LibbyNeural"
|
46 |
+
OUTPUT_FILE = "voice.mp3"
|
47 |
+
#await amain()
|
48 |
+
asyncio.run(amain())
|
49 |
+
return "voice.mp3"
|
50 |
+
|
51 |
+
with gr.Blocks() as demo:
|
52 |
+
gr.HTML("""<h1 align="center">Spelling Tester</h1>""")
|
53 |
+
|
54 |
+
examples = ['SpellingList3.jpg']
|
55 |
+
|
56 |
+
with gr.Row():
|
57 |
+
with gr.Column(scale=1):
|
58 |
+
upload_image = gr.Image(height=400,width=400, value = "SpellingTest8.jpg")
|
59 |
+
#langDrop = gr.Dropdown(
|
60 |
+
# ["ch", "en"], label="Language", info="specify the language will help to increase accuracy.", value = "ch"
|
61 |
+
#),
|
62 |
+
langChoice = gr.Textbox(value="ch", label="Select lanaguage: 'ch' for Chinese, 'en' for English")
|
63 |
+
with gr.Column(scale=3):
|
64 |
+
recognized_text = gr.Textbox(show_label=False, placeholder="spelling list", lines=15)
|
65 |
+
toText = gr.Button("Convert image to text")
|
66 |
+
|
67 |
+
generated_voice = gr.Audio()
|
68 |
+
toVoice = gr.Button("Convert text to voice", variant="primary")
|
69 |
+
|
70 |
+
|
71 |
+
toText.click(
|
72 |
+
image2Text,
|
73 |
+
[upload_image, langChoice],
|
74 |
+
[recognized_text],
|
75 |
+
#show_progress=True,
|
76 |
+
|
77 |
+
)
|
78 |
+
toVoice.click(text2Voice, [recognized_text, langChoice], [generated_voice])
|
79 |
+
|
80 |
+
#searchBtn.click(search, inputs=[user_input], outputs=[search_result], show_progress=True)
|
81 |
+
demo.queue().launch(share=False, inbrowser=True)
|
requirements.txt
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
gradio
|
2 |
+
edge-tts
|
3 |
+
paddlepaddle
|
4 |
+
paddleocr
|