Speech-to-Text-Transcriber-with-GUI

This is a Python-based speech-to-text transcriber application with a user-friendly GUI built using Tkinter. It allows users to record their voice via microphone, convert speech to text using Google’s Speech Recognition API, and view or save the transcribed text. The application is lightweight, beginner-friendly.

Import the necessary libraries

import tkinter as tk from tkinter import messagebox, scrolledtext import speech_recognition as sr

Create the main application window

class SpeechToTextApp: def init(self, root): self.root = root self.root.title("Speech to Text Transcriber") self.root.geometry("500x400") self.root.resizable(False, False)

    # Title label
    self.label = tk.Label(root, text="🎤 Speech to Text", font=("Helvetica", 18, "bold"))
    self.label.pack(pady=10)

    # Transcription display box
    self.text_area = scrolledtext.ScrolledText(root, wrap=tk.WORD, font=("Helvetica", 12), height=10, width=50)
    self.text_area.pack(pady=10)

    # Record button
    self.record_button = tk.Button(root, text="Start Recording", command=self.transcribe_audio, font=("Helvetica", 12), bg="green", fg="white")
    self.record_button.pack(pady=5)

    # Save button (optional)
    self.save_button = tk.Button(root, text="Save Transcription", command=self.save_text, font=("Helvetica", 12), bg="blue", fg="white")
    self.save_button.pack(pady=5)

def transcribe_audio(self):
    recognizer = sr.Recognizer()

    with sr.Microphone() as source:
        try:
            self.text_area.delete(1.0, tk.END)
            self.text_area.insert(tk.END, "Listening...\n")
            self.root.update()

            # Adjusting for ambient noise and setting dynamic energy threshold
            recognizer.adjust_for_ambient_noise(source, duration=1)  # Adjust for 1 second of ambient noise
            self.text_area.insert(tk.END, "Listening for speech...\n")
            self.root.update()

            # Increased duration of listening
            audio = recognizer.listen(source, timeout=None)  # Listen indefinitely until speech ends

            self.text_area.insert(tk.END, "Transcribing...\n")
            self.root.update()

            # Recognize speech using Google Speech Recognition
            text = recognizer.recognize_google(audio)
            self.text_area.delete(1.0, tk.END)
            self.text_area.insert(tk.END, text)

        except sr.WaitTimeoutError:
            messagebox.showerror("Timeout", "No speech detected. Try again.")
        except sr.UnknownValueError:
            messagebox.showerror("Error", "Sorry, could not understand the audio.")
        except sr.RequestError as e:
            messagebox.showerror("API Error", f"Could not request results; {e}")

def save_text(self):
    text = self.text_area.get(1.0, tk.END).strip()
    if text:
        try:
            with open("transcription.txt", "w", encoding='utf-8') as file:
                file.write(text)
            messagebox.showinfo("Saved", "Transcription saved as 'transcription.txt'")
        except Exception as e:
            messagebox.showerror("Save Error", f"Failed to save file: {e}")
    else:
        messagebox.showwarning("No Text", "There's nothing to save!")

Run the app

if name == "main": root = tk.Tk() app = SpeechToTextApp(root) root.mainloop()