File size: 2,426 Bytes
fd15a95 b9eb1b0 fd15a95 11c1d2b fd15a95 |
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 |
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
feature_extractor = AutoFeatureExtractor.from_pretrained("jonatasgrosman/wav2vec2-large-xlsr-53-spanish")
sampling_rate = feature_extractor.sampling_rate
asr = pipeline("automatic-speech-recognition", model="jonatasgrosman/wav2vec2-large-xlsr-53-spanish")
def predict_and_ctc_lm_decode(input_file):
speech = load_and_fix_data(input_file, sampling_rate)
transcribed_text = asr(speech, chunk_length_s=5, stride_length_s=1)["text"]
pipe1 = pipeline("sentiment-analysis", model = "finiteautomata/beto-sentiment-analysis")
sentiment = pipe1(transcribed_text)
sentiment={dic["label"]: dic["score"] for dic in sentiment}
pipe2 = pipeline("text-classification", model = "hackathon-pln-es/twitter_sexismo-finetuned-robertuito-exist2021")
sexism_detection = pipe2(transcribed_text)
sexism_detection={dic["label"]: dic["score"] for dic in sexism_detection}
#sexism_detection = np.where(sexism_detection['label']== 0, 'No Sexista', 'Sexista')
pipe3 = pipeline("text-classification", model = "hackathon-pln-es/twitter_sexismo-finetuned-robertuito-exist2021")
harassment_detection = pipe3(transcribed_text)
harassment_detection={dic["label"]: dic["score"] for dic in harassment_detection}
#harassment_detection = np.where(harassment_detection['label']== 0, 'No Harassment', 'Harassment')
return sexism_detection
#sexism_detection, harassment_detection
gr.Interface(
predict_and_ctc_lm_decode,
inputs=[
gr.inputs.Audio(source="microphone", type="filepath", label="Record your audio")
],
#outputs=[gr.outputs.Label(num_top_classes=2),gr.outputs.Label(num_top_classes=2), gr.outputs.Label(num_top_classes=2)],
outputs=[gr.outputs.Label(num_top_classes=2)],
examples=[["audio1.wav"]],
title="Sentiment Analysis of Spanish Transcribed Audio",
description="This is a Gradio demo for Sentiment Analysis of Transcribed Spanish Audio",
layout="horizontal",
theme="huggingface",
).launch(enable_queue=True, cache_examples=True)
|