frederik
127 lines
3.6 KiB
Python
127 lines
3.6 KiB
Python
import argparse
|
|
import datetime
|
|
import os
|
|
|
|
import numpy as np
|
|
|
|
from utils.helpers import get_settings
|
|
|
|
try:
|
|
import tflite_runtime.interpreter as tflite
|
|
except BaseException:
|
|
from tensorflow import lite as tflite
|
|
|
|
|
|
def loadMetaModel():
|
|
|
|
global M_INTERPRETER
|
|
global M_INPUT_LAYER_INDEX
|
|
global M_OUTPUT_LAYER_INDEX
|
|
global CLASSES
|
|
|
|
# Load TFLite model and allocate tensors.
|
|
M_INTERPRETER = tflite.Interpreter(model_path=userDir + '/BirdNET-Pi/model/BirdNET_GLOBAL_6K_V2.4_MData_Model_FP16.tflite')
|
|
M_INTERPRETER.allocate_tensors()
|
|
|
|
# Get input and output tensors.
|
|
input_details = M_INTERPRETER.get_input_details()
|
|
output_details = M_INTERPRETER.get_output_details()
|
|
|
|
# Get input tensor index
|
|
M_INPUT_LAYER_INDEX = input_details[0]['index']
|
|
M_OUTPUT_LAYER_INDEX = output_details[0]['index']
|
|
|
|
# Load labels
|
|
CLASSES = []
|
|
labelspath = userDir + '/BirdNET-Pi/model/labels.txt'
|
|
with open(labelspath, 'r') as lfile:
|
|
for line in lfile.readlines():
|
|
CLASSES.append(line.replace('\n', ''))
|
|
|
|
print("loaded META model")
|
|
|
|
|
|
def predictFilter(lat, lon, week):
|
|
|
|
global M_INTERPRETER
|
|
|
|
# Does interpreter exist?
|
|
try:
|
|
if M_INTERPRETER is None:
|
|
loadMetaModel()
|
|
except Exception:
|
|
loadMetaModel()
|
|
|
|
# Prepare mdata as sample
|
|
sample = np.expand_dims(np.array([lat, lon, week], dtype='float32'), 0)
|
|
|
|
# Run inference
|
|
M_INTERPRETER.set_tensor(M_INPUT_LAYER_INDEX, sample)
|
|
M_INTERPRETER.invoke()
|
|
|
|
return M_INTERPRETER.get_tensor(M_OUTPUT_LAYER_INDEX)[0]
|
|
|
|
|
|
def explore(lat, lon, week, threshold):
|
|
|
|
# Make filter prediction
|
|
l_filter = predictFilter(lat, lon, week)
|
|
|
|
# Apply threshold
|
|
l_filter = np.where(l_filter >= threshold, l_filter, 0)
|
|
|
|
# Zip with labels
|
|
l_filter = list(zip(l_filter, CLASSES))
|
|
|
|
# Sort by filter value
|
|
l_filter = sorted(l_filter, key=lambda x: x[0], reverse=True)
|
|
|
|
return l_filter
|
|
|
|
|
|
def getSpeciesList(lat, lon, week, threshold=0.05, sort=False):
|
|
|
|
print('Getting species list for {}/{}, Week {}...'.format(lat, lon, week), end='', flush=True)
|
|
|
|
# Extract species from model
|
|
pred = explore(lat, lon, week, threshold)
|
|
|
|
# Make species list
|
|
slist = []
|
|
for p in pred:
|
|
if p[0] >= threshold:
|
|
slist.append([p[1], p[0]])
|
|
|
|
return slist
|
|
|
|
|
|
userDir = os.path.expanduser('~')
|
|
DB_PATH = userDir + '/BirdNET-Pi/scripts/birds.db'
|
|
|
|
conf = get_settings()
|
|
lat = conf.getfloat('LATITUDE')
|
|
lon = conf.getfloat('LONGITUDE')
|
|
|
|
weekofyear = datetime.datetime.today().isocalendar()[1]
|
|
if __name__ == '__main__':
|
|
|
|
# Parse arguments
|
|
parser = argparse.ArgumentParser(
|
|
description='Get list of species for a given location with BirdNET. Sorted by occurrence frequency.'
|
|
)
|
|
parser.add_argument('--threshold', type=float, default=0.05, help='Occurrence frequency threshold. Defaults to 0.05.')
|
|
|
|
args = parser.parse_args()
|
|
|
|
LOCATION_FILTER_THRESHOLD = args.threshold
|
|
|
|
# Get species list
|
|
species_list = getSpeciesList(lat, lon, weekofyear, LOCATION_FILTER_THRESHOLD, False)
|
|
for x in range(len(species_list)):
|
|
print(species_list[x][0] + " - " + str(species_list[x][1]))
|
|
|
|
print("\nThe above species list describes all the species that the model will attempt to detect. \
|
|
If you don't see a species you want detected on this list, decrease your threshold.")
|
|
print("\nNOTE: no actual changes to your BirdNET-Pi species list were made by running this command. \
|
|
To set your desired frequency threshold, do it through the BirdNET-Pi web interface (Tools -> Settings -> Model)")
|