First commit.

app.py

@@ -0,0 +1,81 @@
+import gradio as gr
+from paddleocr import PaddleOCR, draw_ocr
+import asyncio
+import edge_tts
+
+def image2Text(image:str, langChoice:str):
+    ocr = PaddleOCR(use_angle_cls=True, lang=langChoice)  # need to run only once to download and load model into memory
+    img_path = image
+    result = ocr.ocr(img_path, cls=True)
+    text = ""
+    for idx in range(len(result)):
+        res = result[idx]
+        for line in res:
+            import re
+            # remove pinyin if it's Chinese
+            if langChoice=="ch":
+                #t = re.sub('[a-z0-9.]', '', line[1][0])
+                t = re.sub('[a-z]', '', line[1][0])
+                t = re.sub('[0-9]\.', '', t)
+                t = t.replace(" ", "")
+                t = t.replace("（）", "")
+                t = t.replace("（)", "")
+                t = t.replace("( )", "")
+                t = t.replace("()", "")            
+                if t!="":
+                    text +=((t) + "\n")             
+            else:
+                t = line[1][0]
+                t = re.sub('Term [0-9] Spelling', '', t)
+                t = re.sub('Page [0-9]', '', t)
+                if t!="":
+                    text += (t + "\n")
+    return text
+
+    
+def text2Voice(recognized_text, langChoice:str):
+    async def amain() -> None:
+        """Main function"""
+        communicate = edge_tts.Communicate(TEXT, VOICE)
+        await communicate.save(OUTPUT_FILE)
+    TEXT = recognized_text
+    if langChoice == "ch":
+        VOICE = "zh-CN-YunxiaNeural"
+    else:
+        VOICE = "en-GB-LibbyNeural"
+    OUTPUT_FILE = "voice.mp3"
+    #await amain()
+    asyncio.run(amain())
+    return "voice.mp3"
+    
+with gr.Blocks() as demo:
+    gr.HTML("""<h1 align="center">Spelling Tester</h1>""")
+
+    examples = ['SpellingList3.jpg']    
+
+    with gr.Row():
+        with gr.Column(scale=1):
+            upload_image = gr.Image(height=400,width=400, value = "SpellingTest8.jpg")
+            #langDrop = gr.Dropdown(
+            #    ["ch", "en"], label="Language", info="specify the language will help to increase accuracy.", value = "ch"
+            #),            
+            langChoice = gr.Textbox(value="ch", label="Select lanaguage: 'ch' for Chinese, 'en' for English")
+        with gr.Column(scale=3):
+            recognized_text = gr.Textbox(show_label=False, placeholder="spelling list", lines=15)
+    toText = gr.Button("Convert image to text")
+
+    generated_voice = gr.Audio()
+    toVoice = gr.Button("Convert text to voice", variant="primary")
+
+    
+    toText.click(
+        image2Text,
+        [upload_image, langChoice],
+        [recognized_text],
+        #show_progress=True,
+
+    )
+    toVoice.click(text2Voice, [recognized_text, langChoice], [generated_voice])
+
+    #searchBtn.click(search, inputs=[user_input], outputs=[search_result], show_progress=True)
+demo.queue().launch(share=False, inbrowser=True)

requirements.txt

@@ -0,0 +1,4 @@
+gradio
+edge-tts
+paddlepaddle
+paddleocr