Common-Voice-Gender-Detection-IND-En | India (English)

Common-Voice-Gender-Detection-IND-En is a fine-tuned version of facebook/wav2vec2-base-960h for binary audio classification, specifically trained on English (IND-En) speech to detect speaker gender as female or male. This model leverages the Wav2Vec2ForSequenceClassification architecture for efficient and accurate voice-based gender classification.

[!note] Wav2Vec2: Self-Supervised Learning for Speech Recognition: https://arxiv.org/pdf/2006.11477

Classification Report:

              precision    recall  f1-score   support

      Female     0.9986    0.9913    0.9950      3579
        Male     0.9900    0.9984    0.9942      3077

    accuracy                         0.9946      6656
   macro avg     0.9943    0.9949    0.9946      6656
weighted avg     0.9946    0.9946    0.9946      6656

Label Space: 2 Classes

Class 0: female
Class 1: male

Install Dependencies

pip install gradio transformers torch librosa hf_xet

Inference Code

import gradio as gr
from transformers import Wav2Vec2ForSequenceClassification, Wav2Vec2FeatureExtractor
import torch
import librosa

# Load model and processor
model_name = "prithivMLmods/Common-Voice-Gender-Detection-IND-En"
model = Wav2Vec2ForSequenceClassification.from_pretrained(model_name)
processor = Wav2Vec2FeatureExtractor.from_pretrained(model_name)

# Label mapping
id2label = {
    "0": "female",
    "1": "male"
}

def classify_audio(audio_path):
    # Load and resample audio to 16kHz
    speech, sample_rate = librosa.load(audio_path, sr=16000)

    # Process audio
    inputs = processor(
        speech,
        sampling_rate=sample_rate,
        return_tensors="pt",
        padding=True
    )

    with torch.no_grad():
        outputs = model(**inputs)
        logits = outputs.logits
        probs = torch.nn.functional.softmax(logits, dim=1).squeeze().tolist()

    prediction = {
        id2label[str(i)]: round(probs[i], 3) for i in range(len(probs))
    }

    return prediction

# Gradio Interface
iface = gr.Interface(
    fn=classify_audio,
    inputs=gr.Audio(type="filepath", label="Upload Audio (WAV, MP3, etc.)"),
    outputs=gr.Label(num_top_classes=2, label="Gender Classification"),
    title="Common Voice Gender Detection - IND-En",
    description="Upload an English (IND-En) speech clip to classify the speaker's gender as female or male."
)

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

Demo Inference

Intended Use

Common-Voice-Gender-Detection-IND-En is designed for:

• Speech Analytics – Assist in analyzing speaker demographics in call centers or customer service recordings.
• Conversational AI Personalization – Adjust tone or dialogue based on gender detection for more personalized voice assistants.
• Voice Dataset Curation – Automatically tag or filter voice datasets by speaker gender for better dataset management.
• Research Applications – Enable linguistic and acoustic research involving gender-specific speech patterns.
• Multimedia Content Tagging – Automate metadata generation for gender identification in podcasts, interviews, or video content.