Instructions to use Aditya2162/ivus-segmentation with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use Aditya2162/ivus-segmentation with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://Aditya2162/ivus-segmentation") - Notebooks
- Google Colab
- Kaggle
File size: 6,564 Bytes
1d197a4 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 | """Video rendering utilities."""
import numpy as np
from PIL import Image, ImageDraw
from ..processing.preprocessing import ensure_uint8
def draw_contour(draw: ImageDraw.ImageDraw, x_points, y_points, color):
"""Draw a closed contour if points are available."""
if not x_points or not y_points:
return
points = [(float(x_points[i]), float(y_points[i])) for i in range(len(x_points))]
if len(points) < 2:
return
points.append(points[0])
draw.line(points, fill=color, width=2)
def _require_imageio():
try:
import imageio.v2 as imageio
except Exception as exc:
raise RuntimeError(
"Video export requires imageio and imageio-ffmpeg. "
"Install with: python -m pip install imageio imageio-ffmpeg"
) from exc
return imageio
def write_overlay_video(images: np.ndarray, lumen, plaque, video_path: str, fps: float) -> None:
"""Write contour overlay video with lumen/plaque channels."""
imageio = _require_imageio()
images_u8 = ensure_uint8(images)
with imageio.get_writer(video_path, fps=fps) as writer:
for frame_idx in range(images_u8.shape[0]):
frame = Image.fromarray(images_u8[frame_idx], mode="L").convert("RGB")
draw = ImageDraw.Draw(frame)
draw_contour(draw, lumen[0][frame_idx], lumen[1][frame_idx], (0, 255, 0))
draw_contour(draw, plaque[0][frame_idx], plaque[1][frame_idx], (255, 64, 64))
writer.append_data(np.asarray(frame))
def write_overlay_video_with_bifurcation_flags(
images: np.ndarray,
lumen,
plaque,
video_path: str,
fps: float,
bifurcation_probabilities: np.ndarray,
bifurcation_labels: np.ndarray,
threshold: float,
) -> None:
"""Write contour overlay video with per-frame bifurcation text labels."""
imageio = _require_imageio()
images_u8 = ensure_uint8(images)
probs = np.asarray(bifurcation_probabilities, dtype=np.float32)
labels = np.asarray(bifurcation_labels, dtype=np.int32)
with imageio.get_writer(video_path, fps=fps) as writer:
for frame_idx in range(images_u8.shape[0]):
frame = Image.fromarray(images_u8[frame_idx], mode="L").convert("RGB")
draw = ImageDraw.Draw(frame)
draw_contour(draw, lumen[0][frame_idx], lumen[1][frame_idx], (0, 255, 0))
draw_contour(draw, plaque[0][frame_idx], plaque[1][frame_idx], (255, 64, 64))
if frame_idx < probs.shape[0] and frame_idx < labels.shape[0]:
prob = float(probs[frame_idx])
is_bif = bool(labels[frame_idx])
tag = "Branch" if is_bif else "Non-branch"
text = f"{tag} p={prob:.2f} t={float(threshold):.2f}"
text_color = (255, 90, 90) if is_bif else (100, 220, 255)
draw.rectangle([(10, 10), (290, 36)], fill=(0, 0, 0))
draw.text((16, 16), text, fill=text_color)
writer.append_data(np.asarray(frame))
def write_model_comparison_video(images: np.ndarray, tf_lumen, sam_lumen, video_path: str, fps: float) -> None:
"""Write a two-model comparison video (TF red, SAM blue)."""
imageio = _require_imageio()
images_u8 = ensure_uint8(images)
with imageio.get_writer(video_path, fps=fps) as writer:
for frame_idx in range(images_u8.shape[0]):
frame = Image.fromarray(images_u8[frame_idx], mode="L").convert("RGB")
draw = ImageDraw.Draw(frame)
draw_contour(draw, tf_lumen[0][frame_idx], tf_lumen[1][frame_idx], (255, 0, 0))
draw_contour(draw, sam_lumen[0][frame_idx], sam_lumen[1][frame_idx], (0, 0, 255))
writer.append_data(np.asarray(frame))
def _draw_graph_panel(frame_rgb: np.ndarray, scores: np.ndarray, sustained_flags: np.ndarray, frame_idx: int) -> np.ndarray:
panel_h = 150
h, w = frame_rgb.shape[:2]
canvas = Image.new("RGB", (w, h + panel_h), color=(0, 0, 0))
canvas.paste(Image.fromarray(frame_rgb), (0, 0))
draw = ImageDraw.Draw(canvas)
top = h + 12
left = 16
right = w - 16
bottom = h + panel_h - 18
draw.rectangle([(left, top), (right, bottom)], fill=(20, 20, 20), outline=(90, 90, 90), width=1)
n = len(scores)
if n > 1:
in_run = False
run_start = 0
for i in range(n):
if sustained_flags[i] and not in_run:
run_start = i
in_run = True
if (not sustained_flags[i] or i == n - 1) and in_run:
run_end = i if not sustained_flags[i] else i + 1
x0 = left + int((run_start / (n - 1)) * (right - left))
x1 = left + int(((run_end - 1) / (n - 1)) * (right - left))
draw.rectangle([(x0, top), (x1, bottom)], fill=(70, 20, 20))
in_run = False
def px(i):
return left + int((i / (n - 1)) * (right - left))
def py(v):
v = float(np.clip(v, 0.0, 1.0))
return bottom - int(v * (bottom - top))
pts = [(px(i), py(scores[i])) for i in range(n)]
if len(pts) > 1:
draw.line(pts, fill=(120, 255, 120), width=2)
cur = min(max(frame_idx, 0), n - 1)
cx = px(cur)
draw.line([(cx, top), (cx, bottom)], fill=(255, 255, 0), width=2)
flag_txt = "SUSTAINED BRANCH SIGNAL" if sustained_flags[cur] else "normal"
draw.text((left + 6, top + 4), f"Oblongness: {float(scores[cur]):.3f} [{flag_txt}]", fill=(230, 230, 230))
else:
draw.text((left + 6, top + 4), "Oblongness: insufficient data", fill=(230, 230, 230))
return np.asarray(canvas)
def write_overlay_video_with_graph(
images: np.ndarray,
lumen,
plaque,
video_path: str,
fps: float,
oblong_scores: np.ndarray,
sustained_flags: np.ndarray,
) -> None:
"""Write the fused overlay video with an animated bottom graph panel."""
imageio = _require_imageio()
images_u8 = ensure_uint8(images)
with imageio.get_writer(video_path, fps=fps) as writer:
for frame_idx in range(images_u8.shape[0]):
frame = Image.fromarray(images_u8[frame_idx], mode="L").convert("RGB")
draw = ImageDraw.Draw(frame)
draw_contour(draw, lumen[0][frame_idx], lumen[1][frame_idx], (0, 255, 0))
draw_contour(draw, plaque[0][frame_idx], plaque[1][frame_idx], (255, 64, 64))
with_panel = _draw_graph_panel(np.asarray(frame), oblong_scores, sustained_flags, frame_idx)
writer.append_data(with_panel)
|