import gradio as gr
from transformers import pipeline
import numpy as np
import os

accuracy_classifier = pipeline(task="audio-classification", model="JohnJumon/pronunciation_accuracy")
fluency_classifier = pipeline(task="audio-classification", model="JohnJumon/fluency_accuracy")
prosodic_classifier = pipeline(task="audio-classification", model="JohnJumon/prosodic_accuracy")

def pronunciation_scoring(audio):
  accuracy_description = {
      'Extremely Poor': 'Extremely poor pronunciation and only one or two words are recognizable',
      'Poor': 'Poor, clumsy and rigid pronunciation of the sentence as a whole, with serious pronunciation mistakes',
      'Average': 'The overall pronunciation of the sentence is understandable, with many pronunciation mistakes and accent, but it does not affect the understanding of basic meanings',
      'Good': 'The overall pronunciation of the sentence is good, with a few pronunciation mistakes',
      'Excellent': 'The overall pronunciation of the sentence is excellent, with accurate phonology and no obvious pronunciation mistakes'
    }
  fluency_description = {
      'Very Influent': 'Intermittent, very influent speech, with lots of pauses, repetition, and stammering', 
      'Influent': 'The speech is a little influent, with many pauses, repetition, and stammering', 
      'Average': 'Fluent in general, with a few pauses, repetition, and stammering', 
      'Fluent': 'Fluent without noticeable pauses or stammering'
    }
  prosodic_description = {
      'Poor': 'Poor intonation and lots of stammering and pauses, unable to read a complete sentence', 
      'Unstable': 'Unstable speech speed, speak too fast or too slow, without the sense of rhythm', 
      'Stable': 'Unstable speech speed, many stammering and pauses with a poor sense of rhythm', 
      'Almost': 'Nearly correct intonation at a stable speaking speed, nearly smooth and coherent, but with little stammering and few pauses', 
      'Perfect': 'Correct intonation at a stable speaking speed, speak with cadence, and can speak like a native'
    }
  accuracy = accuracy_classifier(audio)
  fluency = fluency_classifier(audio)
  prosodic = prosodic_classifier(audio)
  result = {
      'accuracy': accuracy,
      'fluency': fluency,
      'prosodic': prosodic
      }
  for category, scores in result.items():
    max_score_label = max(scores, key=lambda x: x['score'])['label']
    result[category] = max_score_label
  return result['accuracy'], accuracy_description[result['accuracy']], result['fluency'], fluency_description[result['fluency']], result['prosodic'], prosodic_description[result['prosodic']]

gradio_app = gr.Interface(
    pronunciation_scoring,
    inputs=gr.Audio(sources="microphone", type="filepath"),
    outputs=[
        gr.Label(label="Accuracy Result"),
        gr.Textbox(interactive=False, show_label=False),
        gr.Label(label="Fluency Result"),
        gr.Textbox(interactive=False, show_label=False),
        gr.Label(label="Prosodic Result"),
        gr.Textbox(interactive=False, show_label=False)
      ],
    title="Pronunciation Scoring",
    description="This app will score your pronunciation accuracy, fluency, and prosodic (intonation)",
    examples=[
        [os.path.join(os.path.dirname(__file__),"audio.wav")],
    ]
)

if __name__ == "__main__":
    gradio_app.launch()