|
import gradio as gr |
|
import requests |
|
import json |
|
import os |
|
|
|
def send_req(title, audio_file_path): |
|
url = "https://dev-phonic-api.vuihoc.vn/api/v3/get_score_from_file" |
|
|
|
payload = {'title': title} |
|
files=[ |
|
('audio',('temp.wav',open(audio_file_path,'rb'),'audio/wav')) |
|
] |
|
headers = { |
|
'accept': 'application/json', |
|
'Authorization': 'Bearer eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpZCI6ODg2MSwiZGV2aWNlSWQiOjg4NjEsImlhdCI6MTcxMTk2MTY2OSwiZXhwIjoxNzE3MTQ1NjY5fQ.8Wtzyx3mKVh7K9_GNzseWdK1NH-hycYdNh1uFsoVvEg' |
|
} |
|
|
|
response = requests.request("POST", url, headers=headers, data=payload, files=files) |
|
|
|
return response.json() |
|
|
|
def create_html_output(real_transcript, is_letter_correct_all_words): |
|
html_output = "<center><h1>" |
|
for i, char in enumerate(real_transcript): |
|
if is_letter_correct_all_words[i] == '1': |
|
html_output += f"<span style='color:green;'>{char}</span>" |
|
else: |
|
html_output += f"<span style='color:red;'>{char}</span>" |
|
html_output += "</h1></center>" |
|
return html_output |
|
|
|
def create_html_output_ipa(word_score_list): |
|
html_output = "<center><h1>" |
|
for word_score in word_score_list: |
|
for phone_score in word_score["phone_score_list"]: |
|
if phone_score["quality_score"] == 100: |
|
html_output += f"<span style='color:green;'>{phone_score['phone_ipa']}</span>" |
|
else: |
|
html_output += f"<span style='color:red;'>{phone_score['phone_ipa']}</span>" |
|
html_output += "<span style='color:red;'> </span>" |
|
html_output += "</h1></center>" |
|
return html_output |
|
|
|
def download_audio_file(url, filename=None): |
|
""" |
|
Tải xuống tệp âm thanh từ một URL và lưu nó vào đĩa. |
|
|
|
Args: |
|
url (str): URL của tệp âm thanh cần tải xuống. |
|
filename (str, optional): Tên tệp để lưu (mặc định là tên tệp từ URL). |
|
""" |
|
|
|
|
|
response = requests.get(url) |
|
response.raise_for_status() |
|
|
|
|
|
if not filename: |
|
filename = url.split("/")[-1] |
|
|
|
with open(filename, "wb") as f: |
|
f.write(response.content) |
|
|
|
return filename |
|
|
|
|
|
def pa_check(url_audio, microphone, file_upload, reference_text): |
|
if url_audio: |
|
file = download_audio_file(url_audio) |
|
else: |
|
if (microphone is not None) and (file_upload is not None): |
|
warn_output = ( |
|
"WARNING: You've uploaded an audio file and used the microphone. " |
|
"The recorded file from the microphone will be used and the uploaded audio will be discarded.\n" |
|
) |
|
|
|
elif (microphone is None) and (file_upload is None): |
|
return "ERROR: You have to either use the microphone or upload an audio file" |
|
|
|
file = microphone if microphone is not None else file_upload |
|
|
|
result = send_req(reference_text, file) |
|
try: |
|
|
|
html_output = create_html_output(result["data"]["real_transcripts"], result["data"]["is_letter_correct_all_words"]) |
|
html_output_ipa = create_html_output_ipa(result["data"]["word_score_list"]) |
|
except Exception as e: |
|
print(e) |
|
print(result["data"]["real_transcripts"]) |
|
print(result["data"]["is_letter_correct_all_words"]) |
|
html_output = "ERROR: Something went wrong with the server response. Please try again later." |
|
|
|
|
|
return json.dumps(result, indent=4, ensure_ascii=False), html_output, html_output_ipa |
|
|
|
|
|
demo = gr.Interface( |
|
fn=pa_check, |
|
inputs=[ |
|
gr.Textbox(label="Url audio", type="text", placeholder="Download audio form url"), |
|
gr.Audio(sources="microphone", type="filepath"), |
|
gr.Audio(sources="upload", type="filepath"), |
|
gr.Textbox(label="Reference text", type="text", placeholder="How are you?|What is your name?"), |
|
], |
|
outputs=[ |
|
gr.Textbox(label="Output"), |
|
"html", |
|
"html" |
|
], |
|
theme="huggingface", |
|
title="Pronunciation Assessment", |
|
allow_flagging="never" |
|
) |
|
demo.launch(auth=(os.environ['username'], os.environ['password'])) |
|
|
|
|
|
|
