Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -1,30 +1,23 @@
|
|
1 |
-
import gradio as gr
|
2 |
import matplotlib.pyplot as plt
|
3 |
import numpy as np
|
|
|
|
|
|
|
|
|
|
|
|
|
4 |
import tarfile
|
5 |
import wget
|
|
|
6 |
from huggingface_hub import snapshot_download
|
7 |
import os
|
8 |
-
from object_detection.utils import label_map_util
|
9 |
-
from object_detection.utils import visualization_utils as viz_utils
|
10 |
-
from object_detection.utils import ops as utils_op
|
11 |
-
from six import BytesIO
|
12 |
-
from PIL import Image
|
13 |
-
import tensorflow as tf
|
14 |
-
from huggingface_hub import from_pretrained_keras
|
15 |
-
|
16 |
-
model = from_pretrained_keras("23A066X/23A066X_model")
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
|
21 |
PATH_TO_LABELS = 'label_map.pbtxt'
|
22 |
category_index = label_map_util.create_category_index_from_labelmap(PATH_TO_LABELS, use_display_name=True)
|
23 |
|
24 |
|
25 |
-
|
26 |
def pil_image_as_numpy_array(pilimg):
|
27 |
-
|
28 |
img_array = tf.keras.utils.img_to_array(pilimg)
|
29 |
img_array = np.expand_dims(img_array, axis=0)
|
30 |
return img_array
|
@@ -38,6 +31,7 @@ def load_image_into_numpy_array(path):
|
|
38 |
|
39 |
def load_model():
|
40 |
download_dir = snapshot_download(REPO_ID)
|
|
|
41 |
saved_model_dir = os.path.join(download_dir, "saved_model")
|
42 |
detection_model = tf.saved_model.load(saved_model_dir)
|
43 |
return detection_model
|
@@ -49,6 +43,9 @@ def load_model2():
|
|
49 |
detection_model = tf.saved_model.load(str(model_dir))
|
50 |
return detection_model
|
51 |
|
|
|
|
|
|
|
52 |
|
53 |
def predict(pilimg):
|
54 |
|
@@ -58,10 +55,13 @@ def predict(pilimg):
|
|
58 |
def predict2(image_np):
|
59 |
|
60 |
results = detection_model(image_np)
|
|
|
61 |
# different object detection models have additional results
|
62 |
result = {key:value.numpy() for key,value in results.items()}
|
|
|
63 |
label_id_offset = 0
|
64 |
image_np_with_detections = image_np.copy()
|
|
|
65 |
viz_utils.visualize_boxes_and_labels_on_image_array(
|
66 |
image_np_with_detections[0],
|
67 |
result['detection_boxes'][0],
|
@@ -69,29 +69,32 @@ def predict2(image_np):
|
|
69 |
result['detection_scores'][0],
|
70 |
category_index,
|
71 |
use_normalized_coordinates=True,
|
72 |
-
max_boxes_to_draw=
|
73 |
-
min_score_thresh=0.
|
74 |
agnostic_mode=False,
|
75 |
line_thickness=3)
|
|
|
76 |
result_pil_img = tf.keras.utils.array_to_img(image_np_with_detections[0])
|
77 |
|
78 |
return result_pil_img
|
79 |
|
80 |
|
81 |
-
REPO_ID = "
|
82 |
-
detection_model =
|
83 |
-
# detection_model = load_model()
|
84 |
-
|
85 |
# pil_image = Image.open(image_path)
|
86 |
# image_arr = pil_image_as_numpy_array(pil_image)
|
87 |
|
88 |
# predicted_img = predict(image_arr)
|
89 |
# predicted_img.save('predicted.jpg')
|
90 |
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
title="Cauliflower and Beetroot Detection.",
|
95 |
-
description="Using Model : ssd_resnet50_v1_fpn_640x640_coco17_tpu-8",
|
96 |
-
).launch(share=True)
|
97 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import matplotlib.pyplot as plt
|
2 |
import numpy as np
|
3 |
+
from six import BytesIO
|
4 |
+
from PIL import Image
|
5 |
+
import tensorflow as tf
|
6 |
+
from object_detection.utils import label_map_util
|
7 |
+
from object_detection.utils import visualization_utils as viz_utils
|
8 |
+
from object_detection.utils import ops as utils_op
|
9 |
import tarfile
|
10 |
import wget
|
11 |
+
import gradio as gr
|
12 |
from huggingface_hub import snapshot_download
|
13 |
import os
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
|
15 |
PATH_TO_LABELS = 'label_map.pbtxt'
|
16 |
category_index = label_map_util.create_category_index_from_labelmap(PATH_TO_LABELS, use_display_name=True)
|
17 |
|
18 |
|
|
|
19 |
def pil_image_as_numpy_array(pilimg):
|
20 |
+
|
21 |
img_array = tf.keras.utils.img_to_array(pilimg)
|
22 |
img_array = np.expand_dims(img_array, axis=0)
|
23 |
return img_array
|
|
|
31 |
|
32 |
def load_model():
|
33 |
download_dir = snapshot_download(REPO_ID)
|
34 |
+
# download_dir = os.path.join(download_dir, "saved_model")
|
35 |
saved_model_dir = os.path.join(download_dir, "saved_model")
|
36 |
detection_model = tf.saved_model.load(saved_model_dir)
|
37 |
return detection_model
|
|
|
43 |
detection_model = tf.saved_model.load(str(model_dir))
|
44 |
return detection_model
|
45 |
|
46 |
+
# samples_folder = 'test_samples
|
47 |
+
# image_path = 'test_samples/sample_balloon.jpeg
|
48 |
+
#
|
49 |
|
50 |
def predict(pilimg):
|
51 |
|
|
|
55 |
def predict2(image_np):
|
56 |
|
57 |
results = detection_model(image_np)
|
58 |
+
|
59 |
# different object detection models have additional results
|
60 |
result = {key:value.numpy() for key,value in results.items()}
|
61 |
+
|
62 |
label_id_offset = 0
|
63 |
image_np_with_detections = image_np.copy()
|
64 |
+
|
65 |
viz_utils.visualize_boxes_and_labels_on_image_array(
|
66 |
image_np_with_detections[0],
|
67 |
result['detection_boxes'][0],
|
|
|
69 |
result['detection_scores'][0],
|
70 |
category_index,
|
71 |
use_normalized_coordinates=True,
|
72 |
+
max_boxes_to_draw=200,
|
73 |
+
min_score_thresh=0.60,
|
74 |
agnostic_mode=False,
|
75 |
line_thickness=3)
|
76 |
+
|
77 |
result_pil_img = tf.keras.utils.array_to_img(image_np_with_detections[0])
|
78 |
|
79 |
return result_pil_img
|
80 |
|
81 |
|
82 |
+
REPO_ID = "A23066X/A23066X_model"
|
83 |
+
detection_model = load_model()
|
|
|
|
|
84 |
# pil_image = Image.open(image_path)
|
85 |
# image_arr = pil_image_as_numpy_array(pil_image)
|
86 |
|
87 |
# predicted_img = predict(image_arr)
|
88 |
# predicted_img.save('predicted.jpg')
|
89 |
|
90 |
+
title = "Cauliflower and Beetroot Detection"
|
91 |
+
description = "Using ssd_resnet50_v1_fpn_640x640_coco17_tpu-8"
|
92 |
+
|
|
|
|
|
|
|
93 |
|
94 |
+
gr.Interface(fn=predict,
|
95 |
+
title = title,
|
96 |
+
description = description,
|
97 |
+
css=css_code,
|
98 |
+
inputs=gr.Image(type="pil", height=309),
|
99 |
+
outputs=gr.Image(type="pil", height=350)
|
100 |
+
).launch(share=True)
|