Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| from transformers import AutoProcessor, ASTModel | |
| from tqdm import tqdm | |
| from bing_image_downloader import downloader | |
| import torch | |
| import sklearn | |
| from sklearn.svm import SVC | |
| import numpy as np | |
| from sklearn.model_selection import train_test_split | |
| import pickle | |
| import librosa | |
| import os | |
| import base64 | |
| import matplotlib.pyplot as plt | |
| import webbrowser | |
| import requests | |
| from bs4 import BeautifulSoup | |
| processor = AutoProcessor.from_pretrained("MIT/ast-finetuned-audioset-10-10-0.4593") | |
| model = ASTModel.from_pretrained("MIT/ast-finetuned-audioset-10-10-0.4593") | |
| device = torch.device("cuda" if torch.cuda.is_available() else "cpu") | |
| model.to(device) | |
| model.eval() | |
| background_image = """ | |
| <style> | |
| [data-testid="stAppViewContainer"] > .main { | |
| background-image: url("https://lh3.googleusercontent.com/pw/AP1GczPRGjf9P6yxyTelG-Ku3XLWHg7xeITK2fGZVFAP9zOPChtQuCyZ8jNuLdcjNcHK2gffbYIXtMe9IyMuSJ0PidcLN-A3ou3P1ffvb4aA0LbWIZPuvk6I=w2400"); | |
| background-size: 100vw 100vh; # This sets the size to cover 100% of the viewport width and height | |
| background-position: center; | |
| background-repeat: no-repeat; | |
| } | |
| </style> | |
| """ | |
| st.markdown(background_image, unsafe_allow_html=True) | |
| st.markdown("<h2 style='text-align: center; color: white; font-family: courier ;'>Europe Bird Sound Recognition</h1>", unsafe_allow_html=True) | |
| st.markdown("<h4 style='text-align: left; color: white; font-family: courier; '> Upload Audio Here</h1>", unsafe_allow_html=True) | |
| country_dict = {'Belgium': ['Black Woodpecker', 'African Pied Wagtail'], 'Czech Republic': ['African Pied Wagtail', 'European Turtle Dove', 'River Warbler'], 'Denmark': ['Eurasian Oystercatcher', 'African Pied Wagtail'], 'Estonia': ['African Pied Wagtail', 'Wood Sandpiper'], 'France': ['Black-headed Gull', 'European Nightjar', 'European Herring Gull', 'Rook', 'European Honey Buzzard', 'Western Jackdaw', 'Eurasian Bullfinch', 'African Pied Wagtail', 'Black Woodpecker', 'Common Pheasant'], 'Germany': ['Eurasian Wryneck', 'Eurasian Jay', 'Stock Dove', 'Great Spotted Woodpecker', 'Western Jackdaw', 'Eurasian Oystercatcher', 'African Pied Wagtail', 'Grey Partridge', 'Common Swift', 'Carrion Crow', 'Common Moorhen', 'Northern Raven', 'Eurasian Bullfinch'], 'Italy': ['European Turtle Dove', 'European Honey Buzzard', 'Eurasian Coot', 'African Pied Wagtail'], 'Latvia': ['African Pied Wagtail', 'Wood Sandpiper'], 'Netherlands': ['Eurasian Jay', 'Meadow Pipit', 'Stock Dove', 'Eurasian Coot', 'African Pied Wagtail', 'Carrion Crow', 'Common Moorhen', 'Northern Lapwing'], 'Norway': ['European Herring Gull', 'Meadow Pipit', 'Great Spotted Woodpecker', 'Eurasian Magpie', 'Willow Ptarmigan', 'Wood Sandpiper', 'African Pied Wagtail', 'European Golden Plover'], 'Poland': ['Eurasian Wryneck', 'River Warbler', 'Eurasian Magpie', 'Western Jackdaw', 'Eurasian Bullfinch', 'African Pied Wagtail', 'Northern Raven', 'Common Pheasant'], 'Spain': ['African Pied Wagtail', 'Common Swift'], 'Sweden': ['Eurasian Wryneck', 'European Nightjar', 'Tawny Owl', 'Black Woodpecker', 'Northern Lapwing', 'Northern Raven', 'Common Pheasant', 'Eurasian Oystercatcher', 'Red-throated Loon', 'Common Swift', 'Common Moorhen', 'Great Spotted Woodpecker', 'Eurasian Treecreeper', 'Eurasian Magpie', 'African Pied Wagtail', 'Stock Dove', 'Rook', 'River Warbler', 'Grey Partridge', 'European Golden Plover'], 'United Kingdom': ['Tawny Owl', 'European Honey Buzzard', 'European Greenfinch', 'Eurasian Coot', 'Sedge Warbler', 'Willow Warbler', 'Common Chiffchaff', 'Garden Warbler', 'Corn Bunting', 'Black-headed Gull', 'Common Linnet', 'Eurasian Treecreeper', 'Eurasian Blue Tit', 'Eurasian Wren', 'Common Blackbird', 'African Pied Wagtail', 'European Robin', 'Eurasian Reed Warbler', 'Common Wood Pigeon', 'Song Thrush', 'Barn Swallow', 'Eurasian Skylark', 'Willow Ptarmigan', 'Common Redstart', 'Common Whitethroat', 'Common Nightingale']} | |
| # | |
| option = st.selectbox('Select a country', list(country_dict.keys())) | |
| st.markdown("<h5 style='text-align: center; color: white; font-family: courier; '> Selected country is : "+option+"</h2>", unsafe_allow_html=True) | |
| filename = 'svm_model.sav' | |
| svm_model = pickle.load(open(filename, 'rb')) | |
| uploaded_file = st.file_uploader("Choose a audio file") | |
| frame_len = 22050*2 | |
| label_dict = {0: 'African Pied Wagtail', 1: 'Barn Swallow', 2: 'Black Woodpecker', 3: 'Black-headed Gull', 4: 'Carrion Crow', 5: 'Common Blackbird', 6: 'Common Chiffchaff', 7: 'Common Linnet', 8: 'Common Moorhen', 9: 'Common Nightingale', 10: 'Common Pheasant', 11: 'Common Redstart', 12: 'Common Swift', 13: 'Common Whitethroat', 14: 'Common Wood Pigeon', 15: 'Corn Bunting', 16: 'Eurasian Blue Tit', 17: 'Eurasian Bullfinch', 18: 'Eurasian Coot', 19: 'Eurasian Jay', 20: 'Eurasian Magpie', 21: 'Eurasian Oystercatcher', 22: 'Eurasian Reed Warbler', 23: 'Eurasian Skylark', 24: 'Eurasian Treecreeper', 25: 'Eurasian Wren', 26: 'Eurasian Wryneck', 27: 'European Golden Plover', 28: 'European Greenfinch', 29: 'European Herring Gull', 30: 'European Honey Buzzard', 31: 'European Nightjar', 32: 'European Robin', 33: 'European Turtle Dove', 34: 'Garden Warbler', 35: 'Great Spotted Woodpecker', 36: 'Grey Partridge', 37: 'Meadow Pipit', 38: 'Northern Lapwing', 39: 'Northern Raven', 40: 'Red-throated Loon', 41: 'River Warbler', 42: 'Rook', 43: 'Sedge Warbler', 44: 'Song Thrush', 45: 'Stock Dove', 46: 'Tawny Owl', 47: 'Western Jackdaw', 48: 'Willow Ptarmigan', 49: 'Willow Warbler', 50: 'Wood Sandpiper'} | |
| reversed_dict = {value: key for key, value in label_dict.items()} | |
| data = [] | |
| if uploaded_file is not None: | |
| y, sr = librosa.load(uploaded_file , sr=16000) | |
| y = ((y-np.amin(y))*2)/(np.amax(y) - np.amin(y)) - 1 | |
| org_len = len(y) | |
| intervals = librosa.effects.split(y, top_db=15, ref=np.max) | |
| intervals = intervals.tolist() | |
| y = (y.flatten()).tolist() | |
| nonsilent_y = [] | |
| for p, q in intervals: | |
| nonsilent_y = nonsilent_y + y[p:q + 1] | |
| y = np.array(nonsilent_y) | |
| final_len = len(y) | |
| sil = org_len - final_len | |
| frame_len = 22050 | |
| for j in range(0, len(y), int(frame_len)): | |
| frame = y[j:j + frame_len] | |
| if len(frame) < frame_len: | |
| frame = frame.tolist() + [0] * (frame_len - len(frame)) | |
| frame = np.array(frame) | |
| ast = processor(frame, sampling_rate=sr, return_tensors="pt") | |
| ast = ast.to(device) | |
| with torch.no_grad(): | |
| outputs = model(**ast) | |
| last_hidden_states = outputs.last_hidden_state.cpu().squeeze().mean(axis=0).numpy() | |
| data.append(last_hidden_states) | |
| data = np.array(data) | |
| y_pred = svm_model.decision_function(data) | |
| y_pred_country = country_dict[option] | |
| y_pred_country_label = [reversed_dict[i] for i in y_pred_country] | |
| y_pred_country_logits = [] | |
| for j in range(len(y_pred)): | |
| temp = [] | |
| for i in y_pred_country_label: | |
| temp.append(y_pred[j][i]) | |
| y_pred_country_logits.append(temp) | |
| y_pred_country_logits = np.array(y_pred_country_logits) | |
| y_pred_country_logits_argmax = np.argmax(y_pred_country_logits, axis=1) | |
| max_probabilities = np.max(y_pred_country_logits, axis=1) | |
| class_probabilities = {} | |
| for i, prob in enumerate(max_probabilities): | |
| class_label = y_pred_country_logits_argmax[i] | |
| class_probabilities[class_label] = prob | |
| max_probable_class = max(class_probabilities, key=class_probabilities.get) | |
| prediction = y_pred_country[max_probable_class] | |
| html_code = """ | |
| <div style='text-align: center; background-color: white; color: black; padding: 10px; border-radius: 20px;'> | |
| <h5 style='font-family: courier; color: black;'>Most probable bird in the audio is:</h5> | |
| <h3 style='weight:100 ;font-family: courier; color: black;'>{}</h3> | |
| <p style='font-family: courier; color: black;'>Probability: {:.2f}</p> | |
| </div> | |
| """.format(prediction, class_probabilities[max_probable_class]) | |
| st.markdown(html_code, unsafe_allow_html=True) | |
| #display bird image | |
| img_path = os.path.join(f'data/images/{prediction}/', 'Image_1' + '.jpg') | |
| st.write("") | |
| st.write("") | |
| st.image(img_path, use_column_width=True) | |
| #Extract bird information from wikipedia | |
| bird_name = prediction | |
| bird_name = bird_name.replace(" ", "_") | |
| url = f"https://en.wikipedia.org/wiki/{bird_name}" | |
| page = requests.get(url) | |
| soup = BeautifulSoup(page.content, 'html.parser') | |
| paragraphs = soup.find_all('p') | |
| text = "" | |
| for paragraph in paragraphs: | |
| if len(paragraph.text.split()) > 10: | |
| text += paragraph.text | |
| text = text.split(".") | |
| outText = "" | |
| for i in range(3): | |
| outText += text[i] + "." | |
| st.markdown(f"<h5 style='opacity:.8; background-color: black; padding: 15px; border-radius: 20px; white: black; font-family: courier;'>{outText}</p>", unsafe_allow_html=True) | |