Hugging Face's logo

Spaces:
seba3y
/
CAPT-ReadAloud
Running

App Files Community
CAPT-ReadAloud
File size: 4,905 Bytes 
be1b9b7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
 
import gradio as gr
# from logic import Speaker_speech_analysis
from scipy.io import wavfile
from wav2vec_aligen import speaker_pronunciation_assesment



def create_html_from_scores(word_levels):
    html_output = ''
    for word, level in word_levels:
        if level == '/':
            html_output += f'<span style="color: #0000ff;">{level}</span> '
        elif level == 'Wrong':
          html_output += f'<span style="color: #dc3545;">{word}</span> '
        elif level == 'Understandable':
          html_output += f'<span style="color: #ffc107;">{word}</span> '
        else:
            html_output += f'<span style="color: #28a745;">{word}</span> '
    return html_output
  
def generate_progress_bar(score, label):
    score = round(score, 2)
    score_text = f"{score:.2f}" if score < 100 else "100"
    if score < 30:
        bar_color = "#dc3545" 
    elif score < 60:
        bar_color = "#dc6545" 
    elif score < 80:
        bar_color = "#ffc107"
    else:
        bar_color = "#28a745"
    bar_length = f"{(score / 100) * 100}%"
    return f"""
    <div class="progress-label">{label}:</div>
    <div class="progress-container">
        <div class="progress-bar" style="width: {bar_length}; background-color: {bar_color};">
            <div class="progress-score">{score_text}</div>
        </div>
    </div>
    <div class="progress-max">Max: 100</div>
    """
# CSS to be used in the Gradio Interface




def analyze_audio(text, audio):
# Write the processed audio to a temporary WAV file
    if text is None or audio is None:
      return 'the audio or the text is missing'
    temp_filename = 'temp_audio.wav'
    wavfile.write(temp_filename, audio[0], audio[1])


    result = speaker_pronunciation_assesment(temp_filename)
    accuracy_score = result['pronunciation_accuracy']
    fluency_score   = result['fluency_score']
    total_score       = result['total_score']
    content_scores = result['content_scores']
    
    pronunciation_progress_bar = generate_progress_bar(accuracy_score, "Pronunciation Accuracy")
    fluency_progress_bar = generate_progress_bar(fluency_score, "Fluency Score")
    content_progress_bar = generate_progress_bar(content_scores, "Content Score")
    total_progress_bar = generate_progress_bar(total_score, "Total Score")
    
    
    html_with_css = f"""
    <style>
    .legend {{
      font-size: 22px;
      display: flex;
      align-items: center;
      gap: 12px;
    }}
    
    .legend-dot {{
        height: 15px;
        width: 15px;
        border-radius: 50%;
        display: inline-block;
      }}
      
    .good {{ color: #28a745; 
    }}
    .average {{ color: #ffc107; 
    }}
    .bad {{ color: #dc3545;
    }}
    
    .wrong {{ color: #dc3545;
    }}
        
    .text {{
        font-size: 20px;
        margin-bottom: 20px;
      }}

    .progress-container {{
        width: 100%;
        background-color: #ddd;
        border-radius: 13px;
        overflow: hidden;
      }}

    .progress-bar {{
        height: 30px;
        line-height: 30px;
        text-align: center;
        font-size: 16px;
        border-radius: 15px;
        transition: width 1s ease;
      }}

    .progress-label {{
        font-weight: bold;
        font-size: 22px;
        margin-bottom: 20px;
        margin-top: 5px;
        text-align: center;
      }}

    .progress-score {{
        display: inline-block;
        color: black;
      }}

    .progress-max {{
        text-align: right;
        margin: 10px;
        font-size: 16px;
      }}
        
    </style>
    
    
    <div class="legend">
      <span class="legend-dot" style="background-color: #28a745;"></span><span>Good</span>
      <span class="legend-dot" style="background-color: #ffc107;"></span><span>Understandable</span>
      <span class="legend-dot" style="background-color: #dc3545;"></span><span>Bad</span>
      <span class="legend-dot" style="background-color: #0000ff;"></span><span>No Speech</span>
    </div>
    
    {total_progress_bar}
    {pronunciation_progress_bar}
    {fluency_progress_bar}
    {content_progress_bar}
    """
        # 

    return html_with_css

# Define the Gradio interface
iface = gr.Interface(fn=analyze_audio,
                     inputs=[gr.Textbox(label='Training Text', placeholder='Write the text for pronunciation task', interactive=True, visible=True, show_copy_button=True,), 
                             gr.Audio(label="Recoreded Audio", sources=['microphone', 'upload'])
                             ],
                     outputs=[gr.HTML(label="Analysis of pronunciation"),
                              ],
                    #  css=additional_css,
                     # title="Audio Analysis Tool",
                     description="Write any text and recored an audio to predict pronunciation erors"
                     )

# Run the Gradio app
if __name__ == "__main__":
    iface.launch(share=True)