Spaces:
Runtime error
Runtime error
Upload efficientNet-algo/predict.py with huggingface_hub
Browse files- efficientNet-algo/predict.py +64 -0
efficientNet-algo/predict.py
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
Id=[]
|
| 2 |
+
import numpy as np # linear algebra
|
| 3 |
+
import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)
|
| 4 |
+
import tensorflow as tf
|
| 5 |
+
import matplotlib.pyplot as plt
|
| 6 |
+
from PIL import Image
|
| 7 |
+
import os
|
| 8 |
+
IMAGE_SIZE=256
|
| 9 |
+
BATCH_SIZE=32
|
| 10 |
+
CHANNELS=3
|
| 11 |
+
PATH_NAME='../uploads'
|
| 12 |
+
|
| 13 |
+
classes = [ "cocci" , "healthy" , "ncd" , "salmo" , ]
|
| 14 |
+
|
| 15 |
+
def load_test(path_name):
|
| 16 |
+
|
| 17 |
+
for dirname, _, filenames in os.walk(path_name):
|
| 18 |
+
for filename in filenames:
|
| 19 |
+
Id.append(os.path.join(dirname, filename))
|
| 20 |
+
Id[:5]
|
| 21 |
+
|
| 22 |
+
test=pd.DataFrame()
|
| 23 |
+
test=test.assign(filename=Id)
|
| 24 |
+
test.head()
|
| 25 |
+
return test
|
| 26 |
+
def predict(path_name = '../uploads'):
|
| 27 |
+
PATH_NAME=path_name
|
| 28 |
+
# Assuming model, classes, and test are already defined
|
| 29 |
+
result = []
|
| 30 |
+
test= load_test(path_name)
|
| 31 |
+
# Number of rows and columns in the grid
|
| 32 |
+
num_cols = 3
|
| 33 |
+
num_rows = (len(test.filename) + num_cols - 1) // num_cols
|
| 34 |
+
|
| 35 |
+
# Create a figure with the appropriate size
|
| 36 |
+
plt.figure(figsize=(15, 5 * num_rows))
|
| 37 |
+
|
| 38 |
+
for index, i in enumerate(test.filename):
|
| 39 |
+
img = Image.open(i).convert('RGB')
|
| 40 |
+
img = img.resize((300, 300 * img.size[1] // img.size[0]), Image.Resampling.LANCZOS)
|
| 41 |
+
inp_numpy = np.array(img)[None]
|
| 42 |
+
inp = tf.constant(inp_numpy, dtype='float32')
|
| 43 |
+
|
| 44 |
+
# Get model predictions
|
| 45 |
+
model = tf.saved_model.load('./EfficientNet-model')
|
| 46 |
+
class_scores = model(inp)[0].numpy()
|
| 47 |
+
predicted_class_index = class_scores.argmax()
|
| 48 |
+
|
| 49 |
+
predicted_class = classes[predicted_class_index]
|
| 50 |
+
confidence = class_scores[predicted_class_index]
|
| 51 |
+
|
| 52 |
+
# Append result
|
| 53 |
+
# result.append(predicted_class)
|
| 54 |
+
result.append({"label":predicted_class, "accuracy":confidence})
|
| 55 |
+
|
| 56 |
+
# Plotting the image in the grid
|
| 57 |
+
# plt.subplot(num_rows, num_cols, index + 1)
|
| 58 |
+
# plt.imshow(img)
|
| 59 |
+
# plt.title(f"{predicted_class}\nConfidence: {confidence:.2f}\nFilename: {i}")
|
| 60 |
+
# plt.axis('off')
|
| 61 |
+
return result
|
| 62 |
+
# # Show the grid
|
| 63 |
+
# plt.tight_layout()
|
| 64 |
+
# plt.show()
|