|
|
|
import gradio as gr |
|
import librosa |
|
from transformers import AutoFeatureExtractor, pipeline |
|
|
|
|
|
def load_and_fix_data(input_file, model_sampling_rate): |
|
speech, sample_rate = librosa.load(input_file) |
|
if len(speech.shape) > 1: |
|
speech = speech[:, 0] + speech[:, 1] |
|
if sample_rate != model_sampling_rate: |
|
speech = librosa.resample(speech, sample_rate, model_sampling_rate) |
|
return speech |
|
|
|
|
|
model_name1 = "jonatasgrosman/wav2vec2-xls-r-1b-spanish" |
|
feature_extractor = AutoFeatureExtractor.from_pretrained(model_name1) |
|
sampling_rate = feature_extractor.sampling_rate |
|
asr = pipeline("automatic-speech-recognition", model=model_name1) |
|
|
|
|
|
model_name2 = "hackathon-pln-es/Detect-Acoso-Twitter-Es" |
|
classifier = pipeline("text-classification", model = model_name2) |
|
|
|
|
|
def speech_to_text(input_file): |
|
speech = load_and_fix_data(input_file, sampling_rate) |
|
transcribed_text = asr(speech, chunk_length_s=15, stride_length_s=1)["text"] |
|
return transcribed_text |
|
|
|
|
|
def harassment_detector(transcribed_text): |
|
harassment_detection = classifier(transcribed_text)[0]["label"] |
|
return harassment_detection |
|
|
|
|
|
new_line = "\n\n\n" |
|
def asr_and_harassment_detection(input_file): |
|
transcribed_text = speech_to_text(input_file) |
|
harassment_detection = harassment_detector(transcribed_text) |
|
return f"Audio Transcription :{transcribed_text} {new_line} Audio content is: {harassment_detection}" |
|
|
|
|
|
inputs=[gr.inputs.Audio(source="microphone", type="filepath", label="Record your audio")] |
|
outputs=[gr.outputs.Textbox(label="Predicción")] |
|
examples=[["audio2.wav"], ["sample_audio.wav"], ["test1.wav"], ["test2.wav"]] |
|
title="Spanish-Audio-Transcription-based-Harassment-Detection" |
|
|
|
description = """ This is a Gradio demo for Spanish audio transcription-based harassment detection. To use this, simply provide an audio input (audio recording or via microphone), which will subsequently be transcribed and classified as Harassment/non-harassment pertaining to audio (transcription) with the help of pre-trained models. |
|
|
|
|
|
Pre-trained model used for Spanish ASR: [jonatasgrosman/wav2vec2-xls-r-1b-spanish](https://huggingface.co/jonatasgrosman/wav2vec2-xls-r-1b-spanish) |
|
|
|
|
|
Pre-trained model used for Harassment Detection: [hackathon-pln-es/Detect-Acoso-Twitter-Es](https://huggingface.co/hackathon-pln-es/Detect-Acoso-Twitter-Es)""" |
|
|
|
gr.Interface( |
|
asr_and_harassment_detection, |
|
inputs=inputs, |
|
outputs=outputs, |
|
examples=examples, |
|
title=title, |
|
description=description, |
|
layout="horizontal", |
|
theme="huggingface", |
|
).launch(enable_queue=True) |
|
|