Spaces:
Build error
Build error
update
Browse files
app.py
CHANGED
@@ -25,6 +25,11 @@ def load_lottieurl(url: str):
|
|
25 |
return r.json()
|
26 |
|
27 |
|
|
|
|
|
|
|
|
|
|
|
28 |
lottie_penguin = load_lottieurl('https://assets10.lottiefiles.com/datafiles/Yv8B88Go8kHRZ5T/data.json')
|
29 |
st_lottie(lottie_penguin, height=200)
|
30 |
|
@@ -56,11 +61,10 @@ method = st.sidebar.radio('Choose input source π', options=['Image', 'Webcam'
|
|
56 |
def initial_setup():
|
57 |
df_train = pd.read_csv('full_set.csv')
|
58 |
sub_test_list = sorted(list(df_train['Image'].map(lambda x: get_image(x))))
|
59 |
-
# embeddings = torch.load('embeddings.pt')
|
60 |
with open('embeddings.npy', 'rb') as f:
|
61 |
embeddings = np.load(f)
|
62 |
PATH = 'model_onnx.onnx'
|
63 |
-
ort_session = onnxruntime.InferenceSession(PATH)
|
64 |
input_name = ort_session.get_inputs()[0].name
|
65 |
return df_train, sub_test_list, embeddings, ort_session, input_name
|
66 |
|
|
|
25 |
return r.json()
|
26 |
|
27 |
|
28 |
+
# Configure
|
29 |
+
options = onnxruntime.SessionOptions()
|
30 |
+
options.intra_op_num_threads = 8
|
31 |
+
options.inter_op_num_threads = 8
|
32 |
+
|
33 |
lottie_penguin = load_lottieurl('https://assets10.lottiefiles.com/datafiles/Yv8B88Go8kHRZ5T/data.json')
|
34 |
st_lottie(lottie_penguin, height=200)
|
35 |
|
|
|
61 |
def initial_setup():
|
62 |
df_train = pd.read_csv('full_set.csv')
|
63 |
sub_test_list = sorted(list(df_train['Image'].map(lambda x: get_image(x))))
|
|
|
64 |
with open('embeddings.npy', 'rb') as f:
|
65 |
embeddings = np.load(f)
|
66 |
PATH = 'model_onnx.onnx'
|
67 |
+
ort_session = onnxruntime.InferenceSession(PATH, sess_options=options)
|
68 |
input_name = ort_session.get_inputs()[0].name
|
69 |
return df_train, sub_test_list, embeddings, ort_session, input_name
|
70 |
|
utils.py
CHANGED
@@ -1,6 +1,6 @@
|
|
1 |
-
#import torch
|
2 |
-
#import torch.nn.functional as F
|
3 |
-
#from torchvision import transforms
|
4 |
|
5 |
from PIL import Image
|
6 |
import numpy as np
|
@@ -10,7 +10,7 @@ from numpy.linalg import norm
|
|
10 |
import onnx, os, time, onnxruntime
|
11 |
import pandas as pd
|
12 |
import threading
|
13 |
-
#import queue
|
14 |
import cv2
|
15 |
import av
|
16 |
|
@@ -25,17 +25,17 @@ from streamlit_webrtc import (
|
|
25 |
import args
|
26 |
|
27 |
|
28 |
-
|
29 |
# def to_numpy(tensor):
|
30 |
# return tensor.detach().cpu().numpy() if tensor.requires_grad else tensor.cpu().numpy()
|
31 |
|
|
|
32 |
def get_image(x):
|
33 |
-
|
|
|
34 |
|
35 |
# Transform image to ToTensor
|
36 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
37 |
-
def transform_image(image, IMG
|
38 |
-
|
39 |
# transform = transforms.Compose([
|
40 |
# transforms.Resize((224, 224)),
|
41 |
# transforms.ToTensor(),
|
@@ -45,10 +45,10 @@ def transform_image(image, IMG = True):
|
|
45 |
image = np.asarray(Image.open(image))
|
46 |
# -------------- RESIZE USING CV2 ---------------------
|
47 |
image = cv2.resize(image, dsize=(224, 224))
|
48 |
-
image = np.transpose(image, (2,0,1))
|
49 |
-
#image = (image/255-np.expand_dims(np.array([0.485, 0.456, 0.4065]),axis = (1,2)))/np.expand_dims(np.array([0.229, 0.224, 0.225]),axis = (1,2))
|
50 |
-
image = (image/255-np.array(args.MEAN))/np.array(args.STD)
|
51 |
-
img_transformed = np.expand_dims(image.astype(np.float32), axis
|
52 |
# x = torch.from_numpy(image.astype(np.float32))
|
53 |
# x = torch.transpose(x, 2, 0) # shape [3, 224, 224]
|
54 |
# -------------- RESIZE USING CV2 ---------------------
|
@@ -59,10 +59,10 @@ def transform_image(image, IMG = True):
|
|
59 |
else:
|
60 |
# -------------- RESIZE USING CV2 ---------------------
|
61 |
image = cv2.resize(image, dsize=(224, 224))
|
62 |
-
image = np.transpose(image, (2,0,1))
|
63 |
-
#image = (image/255-np.expand_dims(np.array([0.485, 0.456, 0.4065]),axis = (1,2)))/np.expand_dims(np.array([0.229, 0.224, 0.225]),axis = (1,2))
|
64 |
-
image = (image/255-np.array(args.MEAN))/np.array(args.STD)
|
65 |
-
img_transformed = np.expand_dims(image.astype(np.float32), axis
|
66 |
# x = torch.from_numpy(image.astype(np.float32))
|
67 |
# x = torch.transpose(x, 2, 0)
|
68 |
# -------------- RESIZE USING CV2 ---------------------
|
@@ -72,22 +72,24 @@ def transform_image(image, IMG = True):
|
|
72 |
|
73 |
return img_transformed
|
74 |
|
|
|
75 |
# predict multi-level classification
|
76 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
77 |
def get_classification(image_tensor, df_train, sub_test_list, embeddings,
|
78 |
-
|
79 |
-
):
|
80 |
# Prediction time
|
81 |
start = time.time()
|
82 |
-
#ort_inputs = {input_name: to_numpy(image_tensor)}
|
83 |
ort_inputs = {input_name: image_tensor}
|
84 |
pred, em = ort_session.run(None, ort_inputs)
|
85 |
|
86 |
-
if pred.max(axis=1) > confidence:
|
87 |
# Compute kNN (using Cosine)
|
88 |
-
#knn = torch.nn.CosineSimilarity(dim = 1)(torch.tensor(em), embeddings).topk(1, largest=True)
|
89 |
|
90 |
-
knn = np.array(
|
|
|
91 |
knn = np.argsort(knn)[-1]
|
92 |
|
93 |
# maker = 'Maker: '+str(df_train.iloc[knn.indices.item(), 0])
|
@@ -95,77 +97,81 @@ def get_classification(image_tensor, df_train, sub_test_list, embeddings,
|
|
95 |
# vehicle = str(df_train.iloc[knn.indices.item(), 2])
|
96 |
# year = str(df_train.iloc[knn.indices.item(), 3])
|
97 |
# part = 'Part: '+str(df_train.iloc[knn.indices.item(), 4])
|
98 |
-
maker = 'Maker: '+str(df_train.iloc[knn, 0])
|
99 |
model = str(df_train.iloc[knn, 1])
|
100 |
-
if model=='nan':
|
101 |
-
model='Model: No information'
|
102 |
else:
|
103 |
-
model='Model: '+model
|
104 |
vehicle = str(df_train.iloc[knn, 2])
|
105 |
-
if vehicle=='nan':
|
106 |
-
vehicle='Vehicle: No information'
|
107 |
else:
|
108 |
-
vehicle='Vehicle: '+vehicle
|
109 |
year = str(df_train.iloc[knn, 3])
|
110 |
-
if year=='nan':
|
111 |
-
year='Year: No information'
|
112 |
else:
|
113 |
-
year='Year: '+year
|
114 |
-
part = 'Part: '+str(df_train.iloc[knn, 4])
|
115 |
-
predict_time = 'Predict time: '+str(round(time.time() - start,4))+' seconds'
|
116 |
|
117 |
# Similarity score
|
118 |
-
sim_score = 'Confidence: '+str(round(pred.max(axis=1).item()*100, 2))+'%'
|
119 |
|
120 |
else:
|
121 |
maker = 'This is not car part !'
|
122 |
-
model=vehicle=year=part=predict_time=sim_score=None
|
|
|
|
|
|
|
123 |
|
124 |
-
return {'maker':maker,'model':model,'vehicle':vehicle,'year':year, 'part':part, 'predict_time':predict_time,'sim_score':sim_score}
|
125 |
|
126 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
127 |
def get_classification_frame(image_tensor, df_train, sub_test_list, embeddings,
|
128 |
-
|
129 |
-
):
|
130 |
-
|
131 |
-
#ort_inputs = {input_name: to_numpy(image_tensor)}
|
132 |
ort_inputs = {input_name: image_tensor}
|
133 |
pred, em = ort_session.run(None, ort_inputs)
|
134 |
|
135 |
if pred.max(axis=1) > args.VIDEO_CONFIDENCE:
|
136 |
# knn = torch.nn.CosineSimilarity(dim = 1)(torch.tensor(em), embeddings).topk(1, largest=True)
|
137 |
# part = str(df_train.iloc[knn.indices.item(), 4])
|
138 |
-
knn = np.array(
|
|
|
139 |
knn = np.argsort(knn)[-1]
|
140 |
part = str(df_train.iloc[knn, 4])
|
141 |
# Similarity score
|
142 |
-
sim_score = str(round(pred.max(axis=1).item()*100, 2))+'%'
|
143 |
else:
|
144 |
part = 'No part detected'
|
145 |
sim_score = ''
|
146 |
|
147 |
-
return {'part_name':part,'sim_score':sim_score}
|
|
|
148 |
|
149 |
# predict similarity
|
150 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
151 |
def get_similarity(image_tensor, df_train, sub_test_list, embeddings,
|
152 |
-
|
153 |
-
):
|
154 |
start = time.time()
|
155 |
-
#ort_inputs = {input_name: to_numpy(image_tensor)}
|
156 |
ort_inputs = {input_name: image_tensor}
|
157 |
pred, em = ort_session.run(None, ort_inputs)
|
158 |
|
159 |
# Compute kNN (using Cosine)
|
160 |
-
#knn = torch.nn.CosineSimilarity(dim = 1)(torch.tensor(em), embeddings).topk(6, largest=True)
|
161 |
# idx = knn.indices.numpy()
|
162 |
-
knn = np.array(
|
|
|
163 |
idx = np.argsort(knn)[-6:]
|
164 |
-
predict_time = 'Predict time: '+str(round(time.time() - start,4))+' seconds'
|
165 |
images_path = 'Test_set'
|
166 |
images = [os.path.join(images_path, sub_test_list[i]) for i in idx]
|
167 |
# sub_test_list
|
168 |
-
return {'images': images, 'predict_time':predict_time}
|
169 |
|
170 |
|
171 |
# --------------------------------------------------------------------------------------------
|
@@ -173,13 +179,12 @@ def get_similarity(image_tensor, df_train, sub_test_list, embeddings,
|
|
173 |
# --------------------------------------------------------------------------------------------
|
174 |
|
175 |
content_images_dict = {
|
176 |
-
name: os.path.join(args.IMAGES_PATH, filee) for name, filee in
|
|
|
177 |
}
|
178 |
|
179 |
|
180 |
-
|
181 |
def show_original():
|
182 |
-
|
183 |
""" Show Uploaded or Example image before prediction
|
184 |
|
185 |
Returns:
|
@@ -188,7 +193,7 @@ def show_original():
|
|
188 |
path to image
|
189 |
"""
|
190 |
|
191 |
-
if st.sidebar.checkbox('Upload', value=
|
192 |
content_file = st.sidebar.file_uploader("", type=["png", "jpg", "jpeg"])
|
193 |
else:
|
194 |
content_name = st.sidebar.selectbox("or Choose an example Image below", args.CONTENT_IMAGES_NAME)
|
@@ -196,7 +201,7 @@ def show_original():
|
|
196 |
|
197 |
col1, col2 = st.columns(2)
|
198 |
with col1:
|
199 |
-
#col1.markdown('## Target image')
|
200 |
if content_file:
|
201 |
col1.write('')
|
202 |
col1.image(content_file, channels='BGR', width=300, clamp=True, caption='Input image')
|
@@ -205,7 +210,6 @@ def show_original():
|
|
205 |
|
206 |
|
207 |
def image_input(content_file, df_train, sub_test_list, embeddings, ort_session, input_name, col2):
|
208 |
-
|
209 |
# Set confidence level
|
210 |
confidence_threshold = st.slider(
|
211 |
"Confidence threshold", 0.0, 1.0, args.DEFAULT_CONFIDENCE_THRESHOLD, 0.05,
|
@@ -232,9 +236,9 @@ def image_input(content_file, df_train, sub_test_list, embeddings, ort_session,
|
|
232 |
if col7.button("SEARCH SIMILAR"):
|
233 |
print_classification(col2, content_file, pred_info)
|
234 |
|
235 |
-
if pred_info['maker']!='This is not car part !':
|
236 |
-
#container = st.container()
|
237 |
-
print_similar_img(pred_images)
|
238 |
else:
|
239 |
st.warning("No similar car part image ! Reduce confidence threshold OR Choose another image.")
|
240 |
else:
|
@@ -278,7 +282,7 @@ def webcam_input(df_train, sub_test_list, embeddings, ort_session, input_name):
|
|
278 |
|
279 |
def recv(self, frame: av.VideoFrame) -> av.VideoFrame:
|
280 |
image = frame.to_ndarray(format="bgr24")
|
281 |
-
content = transform_image(image, IMG
|
282 |
pred_info = get_classification_frame(
|
283 |
content, df_train, sub_test_list,
|
284 |
embeddings, ort_session, input_name
|
@@ -297,14 +301,13 @@ def webcam_input(df_train, sub_test_list, embeddings, ort_session, input_name):
|
|
297 |
|
298 |
|
299 |
def print_classification(col2, content_file, pred_info):
|
300 |
-
|
301 |
""" Print classification prediction
|
302 |
"""
|
303 |
|
304 |
with col2:
|
305 |
col2.markdown('### Predicted information')
|
306 |
col2.markdown('')
|
307 |
-
if pred_info['maker']!='This is not car part !':
|
308 |
col2.markdown('#### - {}'.format(pred_info['maker']))
|
309 |
col2.markdown('#### - {}'.format(pred_info['model']))
|
310 |
col2.markdown('#### - {}'.format(pred_info['vehicle']))
|
@@ -315,8 +318,8 @@ def print_classification(col2, content_file, pred_info):
|
|
315 |
else:
|
316 |
col2.markdown('### {}'.format(pred_info['maker']))
|
317 |
|
318 |
-
def print_similar_img(pred_images):
|
319 |
|
|
|
320 |
""" Print similarity images prediction
|
321 |
"""
|
322 |
|
@@ -325,13 +328,13 @@ def print_similar_img(pred_images):
|
|
325 |
|
326 |
col3, col4, col5 = st.columns(3)
|
327 |
with col3:
|
328 |
-
col3.image(pred_images['images'][0], channels='BGR', clamp=True, width
|
329 |
-
col3.image(pred_images['images'][1], channels='BGR', clamp=True, width
|
330 |
|
331 |
with col4:
|
332 |
-
#col4.markdown('# ')
|
333 |
-
col4.image(pred_images['images'][3], channels='BGR', clamp=True, width
|
334 |
-
col4.image(pred_images['images'][4], channels='BGR', clamp=True, width
|
335 |
with col5:
|
336 |
-
col5.image(pred_images['images'][5], channels='BGR', clamp=True, width
|
337 |
-
col5.image(pred_images['images'][2], channels='BGR', clamp=True, width
|
|
|
1 |
+
# import torch
|
2 |
+
# import torch.nn.functional as F
|
3 |
+
# from torchvision import transforms
|
4 |
|
5 |
from PIL import Image
|
6 |
import numpy as np
|
|
|
10 |
import onnx, os, time, onnxruntime
|
11 |
import pandas as pd
|
12 |
import threading
|
13 |
+
# import queue
|
14 |
import cv2
|
15 |
import av
|
16 |
|
|
|
25 |
import args
|
26 |
|
27 |
|
|
|
28 |
# def to_numpy(tensor):
|
29 |
# return tensor.detach().cpu().numpy() if tensor.requires_grad else tensor.cpu().numpy()
|
30 |
|
31 |
+
|
32 |
def get_image(x):
|
33 |
+
return x.split(', ')[0]
|
34 |
+
|
35 |
|
36 |
# Transform image to ToTensor
|
37 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
38 |
+
def transform_image(image, IMG=True):
|
|
|
39 |
# transform = transforms.Compose([
|
40 |
# transforms.Resize((224, 224)),
|
41 |
# transforms.ToTensor(),
|
|
|
45 |
image = np.asarray(Image.open(image))
|
46 |
# -------------- RESIZE USING CV2 ---------------------
|
47 |
image = cv2.resize(image, dsize=(224, 224))
|
48 |
+
image = np.transpose(image, (2, 0, 1))
|
49 |
+
# image = (image/255-np.expand_dims(np.array([0.485, 0.456, 0.4065]),axis = (1,2)))/np.expand_dims(np.array([0.229, 0.224, 0.225]),axis = (1,2))
|
50 |
+
image = (image / 255 - np.array(args.MEAN)) / np.array(args.STD)
|
51 |
+
img_transformed = np.expand_dims(image.astype(np.float32), axis=0)
|
52 |
# x = torch.from_numpy(image.astype(np.float32))
|
53 |
# x = torch.transpose(x, 2, 0) # shape [3, 224, 224]
|
54 |
# -------------- RESIZE USING CV2 ---------------------
|
|
|
59 |
else:
|
60 |
# -------------- RESIZE USING CV2 ---------------------
|
61 |
image = cv2.resize(image, dsize=(224, 224))
|
62 |
+
image = np.transpose(image, (2, 0, 1))
|
63 |
+
# image = (image/255-np.expand_dims(np.array([0.485, 0.456, 0.4065]),axis = (1,2)))/np.expand_dims(np.array([0.229, 0.224, 0.225]),axis = (1,2))
|
64 |
+
image = (image / 255 - np.array(args.MEAN)) / np.array(args.STD)
|
65 |
+
img_transformed = np.expand_dims(image.astype(np.float32), axis=0)
|
66 |
# x = torch.from_numpy(image.astype(np.float32))
|
67 |
# x = torch.transpose(x, 2, 0)
|
68 |
# -------------- RESIZE USING CV2 ---------------------
|
|
|
72 |
|
73 |
return img_transformed
|
74 |
|
75 |
+
|
76 |
# predict multi-level classification
|
77 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
78 |
def get_classification(image_tensor, df_train, sub_test_list, embeddings,
|
79 |
+
ort_session, input_name, confidence
|
80 |
+
):
|
81 |
# Prediction time
|
82 |
start = time.time()
|
83 |
+
# ort_inputs = {input_name: to_numpy(image_tensor)}
|
84 |
ort_inputs = {input_name: image_tensor}
|
85 |
pred, em = ort_session.run(None, ort_inputs)
|
86 |
|
87 |
+
if pred.max(axis=1) > confidence: # threshold to select of item is car part or not, Yes if > 0.5
|
88 |
# Compute kNN (using Cosine)
|
89 |
+
# knn = torch.nn.CosineSimilarity(dim = 1)(torch.tensor(em), embeddings).topk(1, largest=True)
|
90 |
|
91 |
+
knn = np.array(
|
92 |
+
[dot((em), embeddings[i]) / (norm(em) * norm(embeddings[i])) for i in range(embeddings.shape[0])]).flatten()
|
93 |
knn = np.argsort(knn)[-1]
|
94 |
|
95 |
# maker = 'Maker: '+str(df_train.iloc[knn.indices.item(), 0])
|
|
|
97 |
# vehicle = str(df_train.iloc[knn.indices.item(), 2])
|
98 |
# year = str(df_train.iloc[knn.indices.item(), 3])
|
99 |
# part = 'Part: '+str(df_train.iloc[knn.indices.item(), 4])
|
100 |
+
maker = 'Maker: ' + str(df_train.iloc[knn, 0])
|
101 |
model = str(df_train.iloc[knn, 1])
|
102 |
+
if model == 'nan':
|
103 |
+
model = 'Model: No information'
|
104 |
else:
|
105 |
+
model = 'Model: ' + model
|
106 |
vehicle = str(df_train.iloc[knn, 2])
|
107 |
+
if vehicle == 'nan':
|
108 |
+
vehicle = 'Vehicle: No information'
|
109 |
else:
|
110 |
+
vehicle = 'Vehicle: ' + vehicle
|
111 |
year = str(df_train.iloc[knn, 3])
|
112 |
+
if year == 'nan':
|
113 |
+
year = 'Year: No information'
|
114 |
else:
|
115 |
+
year = 'Year: ' + year
|
116 |
+
part = 'Part: ' + str(df_train.iloc[knn, 4])
|
117 |
+
predict_time = 'Predict time: ' + str(round(time.time() - start, 4)) + ' seconds'
|
118 |
|
119 |
# Similarity score
|
120 |
+
sim_score = 'Confidence: ' + str(round(pred.max(axis=1).item() * 100, 2)) + '%'
|
121 |
|
122 |
else:
|
123 |
maker = 'This is not car part !'
|
124 |
+
model = vehicle = year = part = predict_time = sim_score = None
|
125 |
+
|
126 |
+
return {'maker': maker, 'model': model, 'vehicle': vehicle, 'year': year, 'part': part,
|
127 |
+
'predict_time': predict_time, 'sim_score': sim_score}
|
128 |
|
|
|
129 |
|
130 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
131 |
def get_classification_frame(image_tensor, df_train, sub_test_list, embeddings,
|
132 |
+
ort_session, input_name
|
133 |
+
):
|
134 |
+
# ort_inputs = {input_name: to_numpy(image_tensor)}
|
|
|
135 |
ort_inputs = {input_name: image_tensor}
|
136 |
pred, em = ort_session.run(None, ort_inputs)
|
137 |
|
138 |
if pred.max(axis=1) > args.VIDEO_CONFIDENCE:
|
139 |
# knn = torch.nn.CosineSimilarity(dim = 1)(torch.tensor(em), embeddings).topk(1, largest=True)
|
140 |
# part = str(df_train.iloc[knn.indices.item(), 4])
|
141 |
+
knn = np.array(
|
142 |
+
[dot((em), embeddings[i]) / (norm(em) * norm(embeddings[i])) for i in range(embeddings.shape[0])]).flatten()
|
143 |
knn = np.argsort(knn)[-1]
|
144 |
part = str(df_train.iloc[knn, 4])
|
145 |
# Similarity score
|
146 |
+
sim_score = str(round(pred.max(axis=1).item() * 100, 2)) + '%'
|
147 |
else:
|
148 |
part = 'No part detected'
|
149 |
sim_score = ''
|
150 |
|
151 |
+
return {'part_name': part, 'sim_score': sim_score}
|
152 |
+
|
153 |
|
154 |
# predict similarity
|
155 |
@st.cache(suppress_st_warning=True, allow_output_mutation=True)
|
156 |
def get_similarity(image_tensor, df_train, sub_test_list, embeddings,
|
157 |
+
ort_session, input_name
|
158 |
+
):
|
159 |
start = time.time()
|
160 |
+
# ort_inputs = {input_name: to_numpy(image_tensor)}
|
161 |
ort_inputs = {input_name: image_tensor}
|
162 |
pred, em = ort_session.run(None, ort_inputs)
|
163 |
|
164 |
# Compute kNN (using Cosine)
|
165 |
+
# knn = torch.nn.CosineSimilarity(dim = 1)(torch.tensor(em), embeddings).topk(6, largest=True)
|
166 |
# idx = knn.indices.numpy()
|
167 |
+
knn = np.array(
|
168 |
+
[dot((em), embeddings[i]) / (norm(em) * norm(embeddings[i])) for i in range(embeddings.shape[0])]).flatten()
|
169 |
idx = np.argsort(knn)[-6:]
|
170 |
+
predict_time = 'Predict time: ' + str(round(time.time() - start, 4)) + ' seconds'
|
171 |
images_path = 'Test_set'
|
172 |
images = [os.path.join(images_path, sub_test_list[i]) for i in idx]
|
173 |
# sub_test_list
|
174 |
+
return {'images': images, 'predict_time': predict_time}
|
175 |
|
176 |
|
177 |
# --------------------------------------------------------------------------------------------
|
|
|
179 |
# --------------------------------------------------------------------------------------------
|
180 |
|
181 |
content_images_dict = {
|
182 |
+
name: os.path.join(args.IMAGES_PATH, filee) for name, filee in
|
183 |
+
zip(args.CONTENT_IMAGES_NAME, args.CONTENT_IMAGES_FILE)
|
184 |
}
|
185 |
|
186 |
|
|
|
187 |
def show_original():
|
|
|
188 |
""" Show Uploaded or Example image before prediction
|
189 |
|
190 |
Returns:
|
|
|
193 |
path to image
|
194 |
"""
|
195 |
|
196 |
+
if st.sidebar.checkbox('Upload', value=True, help='Select Upload to browse image from local machine'):
|
197 |
content_file = st.sidebar.file_uploader("", type=["png", "jpg", "jpeg"])
|
198 |
else:
|
199 |
content_name = st.sidebar.selectbox("or Choose an example Image below", args.CONTENT_IMAGES_NAME)
|
|
|
201 |
|
202 |
col1, col2 = st.columns(2)
|
203 |
with col1:
|
204 |
+
# col1.markdown('## Target image')
|
205 |
if content_file:
|
206 |
col1.write('')
|
207 |
col1.image(content_file, channels='BGR', width=300, clamp=True, caption='Input image')
|
|
|
210 |
|
211 |
|
212 |
def image_input(content_file, df_train, sub_test_list, embeddings, ort_session, input_name, col2):
|
|
|
213 |
# Set confidence level
|
214 |
confidence_threshold = st.slider(
|
215 |
"Confidence threshold", 0.0, 1.0, args.DEFAULT_CONFIDENCE_THRESHOLD, 0.05,
|
|
|
236 |
if col7.button("SEARCH SIMILAR"):
|
237 |
print_classification(col2, content_file, pred_info)
|
238 |
|
239 |
+
if pred_info['maker'] != 'This is not car part !':
|
240 |
+
# container = st.container()
|
241 |
+
print_similar_img(pred_images) # , container)
|
242 |
else:
|
243 |
st.warning("No similar car part image ! Reduce confidence threshold OR Choose another image.")
|
244 |
else:
|
|
|
282 |
|
283 |
def recv(self, frame: av.VideoFrame) -> av.VideoFrame:
|
284 |
image = frame.to_ndarray(format="bgr24")
|
285 |
+
content = transform_image(image, IMG=False)
|
286 |
pred_info = get_classification_frame(
|
287 |
content, df_train, sub_test_list,
|
288 |
embeddings, ort_session, input_name
|
|
|
301 |
|
302 |
|
303 |
def print_classification(col2, content_file, pred_info):
|
|
|
304 |
""" Print classification prediction
|
305 |
"""
|
306 |
|
307 |
with col2:
|
308 |
col2.markdown('### Predicted information')
|
309 |
col2.markdown('')
|
310 |
+
if pred_info['maker'] != 'This is not car part !':
|
311 |
col2.markdown('#### - {}'.format(pred_info['maker']))
|
312 |
col2.markdown('#### - {}'.format(pred_info['model']))
|
313 |
col2.markdown('#### - {}'.format(pred_info['vehicle']))
|
|
|
318 |
else:
|
319 |
col2.markdown('### {}'.format(pred_info['maker']))
|
320 |
|
|
|
321 |
|
322 |
+
def print_similar_img(pred_images):
|
323 |
""" Print similarity images prediction
|
324 |
"""
|
325 |
|
|
|
328 |
|
329 |
col3, col4, col5 = st.columns(3)
|
330 |
with col3:
|
331 |
+
col3.image(pred_images['images'][0], channels='BGR', clamp=True, width=300)
|
332 |
+
col3.image(pred_images['images'][1], channels='BGR', clamp=True, width=300)
|
333 |
|
334 |
with col4:
|
335 |
+
# col4.markdown('# ')
|
336 |
+
col4.image(pred_images['images'][3], channels='BGR', clamp=True, width=300)
|
337 |
+
col4.image(pred_images['images'][4], channels='BGR', clamp=True, width=300)
|
338 |
with col5:
|
339 |
+
col5.image(pred_images['images'][5], channels='BGR', clamp=True, width=300)
|
340 |
+
col5.image(pred_images['images'][2], channels='BGR', clamp=True, width=300)
|