Spaces:
Runtime error
Runtime error
feat: add medsam
Browse files- app.py +25 -8
- models/{sam_vit_h_4b8939.pth → medsam_vitb.pth} +2 -2
- models/sam_vit_b_01ec64.pth +3 -0
- requirements.txt +1 -0
- samples/breast_cancer.png +0 -0
- scripts/example.py +1 -0
app.py
CHANGED
@@ -5,12 +5,21 @@ import torch
|
|
5 |
import cv2
|
6 |
from segment_anything import SamPredictor, sam_model_registry
|
7 |
|
8 |
-
|
9 |
-
|
|
|
|
|
10 |
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
|
11 |
-
|
|
|
|
|
|
|
12 |
SAM.to(device=DEVICE)
|
13 |
SAM_PREDICTOR = SamPredictor(SAM)
|
|
|
|
|
|
|
|
|
14 |
|
15 |
|
16 |
def draw_contour(image: np.ndarray, mask: np.ndarray) -> np.ndarray:
|
@@ -23,17 +32,20 @@ def draw_contour(image: np.ndarray, mask: np.ndarray) -> np.ndarray:
|
|
23 |
return contour_image, contours
|
24 |
|
25 |
|
26 |
-
def inference(
|
|
|
|
|
27 |
"""Inference."""
|
28 |
-
|
29 |
|
30 |
input_point = np.array([[coord_y, coord_x]])
|
31 |
input_label = np.array([1])
|
32 |
-
mask, _, _ =
|
33 |
point_coords=input_point,
|
34 |
point_labels=input_label,
|
35 |
multimask_output=False,
|
36 |
)
|
|
|
37 |
h, w = mask.shape[-2:]
|
38 |
mask = mask.reshape(h, w, 1)
|
39 |
mask = (mask * 255).astype(np.uint8)
|
@@ -63,7 +75,7 @@ with gr.Blocks() as demo:
|
|
63 |
)
|
64 |
|
65 |
# Segment image
|
66 |
-
with gr.Tab(label="
|
67 |
with gr.Row().style(equal_height=True):
|
68 |
with gr.Column(label="Input Image"):
|
69 |
# input image
|
@@ -80,7 +92,12 @@ with gr.Blocks() as demo:
|
|
80 |
input_image.select(get_coords, None, [coord_h, coord_w])
|
81 |
gr.Examples(
|
82 |
examples=[
|
83 |
-
[os.path.join(os.path.dirname(__file__), "samples/bears.jpg"), 1300, 950]
|
|
|
|
|
|
|
|
|
|
|
84 |
],
|
85 |
inputs=[input_image, coord_h, coord_w],
|
86 |
outputs=output,
|
|
|
5 |
import cv2
|
6 |
from segment_anything import SamPredictor, sam_model_registry
|
7 |
|
8 |
+
# Global variables
|
9 |
+
OFFICIAL_CHECKPOINT = "./models/sam_vit_b_01ec64.pth"
|
10 |
+
MEDSAM_CHECKPOINT = "./models/medsam_vitb.pth"
|
11 |
+
MODEL_TYPE = "vit_b"
|
12 |
DEVICE = "cuda" if torch.cuda.is_available() else "cpu"
|
13 |
+
|
14 |
+
# Model
|
15 |
+
## OFFICIAL SAM
|
16 |
+
SAM = sam_model_registry[MODEL_TYPE](checkpoint=OFFICIAL_CHECKPOINT)
|
17 |
SAM.to(device=DEVICE)
|
18 |
SAM_PREDICTOR = SamPredictor(SAM)
|
19 |
+
## MEDSAM
|
20 |
+
MEDSAM = sam_model_registry[MODEL_TYPE](checkpoint=MEDSAM_CHECKPOINT)
|
21 |
+
MEDSAM.to(device=DEVICE)
|
22 |
+
MEDSAM_PREDICTOR = SamPredictor(MEDSAM)
|
23 |
|
24 |
|
25 |
def draw_contour(image: np.ndarray, mask: np.ndarray) -> np.ndarray:
|
|
|
32 |
return contour_image, contours
|
33 |
|
34 |
|
35 |
+
def inference(
|
36 |
+
predictor: SamPredictor, image: np.ndarray, coord_y: int, coord_x: int
|
37 |
+
) -> np.ndarray:
|
38 |
"""Inference."""
|
39 |
+
predictor.set_image(image)
|
40 |
|
41 |
input_point = np.array([[coord_y, coord_x]])
|
42 |
input_label = np.array([1])
|
43 |
+
mask, _, _ = predictor.predict(
|
44 |
point_coords=input_point,
|
45 |
point_labels=input_label,
|
46 |
multimask_output=False,
|
47 |
)
|
48 |
+
|
49 |
h, w = mask.shape[-2:]
|
50 |
mask = mask.reshape(h, w, 1)
|
51 |
mask = (mask * 255).astype(np.uint8)
|
|
|
75 |
)
|
76 |
|
77 |
# Segment image
|
78 |
+
with gr.Tab(label="SAM Inference"):
|
79 |
with gr.Row().style(equal_height=True):
|
80 |
with gr.Column(label="Input Image"):
|
81 |
# input image
|
|
|
92 |
input_image.select(get_coords, None, [coord_h, coord_w])
|
93 |
gr.Examples(
|
94 |
examples=[
|
95 |
+
[os.path.join(os.path.dirname(__file__), "samples/bears.jpg"), 1300, 950],
|
96 |
+
[
|
97 |
+
os.path.join(os.path.dirname(__file__), "samples/breast_cancer.png"),
|
98 |
+
125,
|
99 |
+
60,
|
100 |
+
],
|
101 |
],
|
102 |
inputs=[input_image, coord_h, coord_w],
|
103 |
outputs=output,
|
models/{sam_vit_h_4b8939.pth → medsam_vitb.pth}
RENAMED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ec2df62732614e57411cdcf32a23ffdf28910380d03139ee0f4fcbe91eb8c912
|
3 |
+
size 375042383
|
models/sam_vit_b_01ec64.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ec2df62732614e57411cdcf32a23ffdf28910380d03139ee0f4fcbe91eb8c912
|
3 |
+
size 375042383
|
requirements.txt
CHANGED
@@ -1,5 +1,6 @@
|
|
1 |
opencv-python
|
2 |
matplotlib
|
3 |
gradio
|
|
|
4 |
torch
|
5 |
torchvision
|
|
|
1 |
opencv-python
|
2 |
matplotlib
|
3 |
gradio
|
4 |
+
transformers
|
5 |
torch
|
6 |
torchvision
|
samples/breast_cancer.png
ADDED
![]() |
scripts/example.py
CHANGED
@@ -16,6 +16,7 @@ device = "cpu"
|
|
16 |
# Load model
|
17 |
sam = sam_model_registry[model_type](checkpoint=sam_checkpoint)
|
18 |
sam.to(device=device)
|
|
|
19 |
predictor = SamPredictor(sam)
|
20 |
|
21 |
# Preprocessing the image
|
|
|
16 |
# Load model
|
17 |
sam = sam_model_registry[model_type](checkpoint=sam_checkpoint)
|
18 |
sam.to(device=device)
|
19 |
+
|
20 |
predictor = SamPredictor(sam)
|
21 |
|
22 |
# Preprocessing the image
|