tutorials/lodestar-A

[Pwhsky]

updated lodestar tutorial A

[edudc]

There's a small typo under 4. Training LodeSTAR that says "Initialie". Otherwise lgm.

[foca13]

Suggestions

• Specify that this example contains 10 images:
  In mardown cell

1. Loading the Images of Quantum Dots
   The data is found in the tutorials/2-examples/assets directory as an array with the format (time, height, width). This example contains 10 imatges
   Then
   images = np.load("./assets/image_quantum_dots.npy")
   images.shape

• Plot the final predictions in a single 2x5 plot, something like:
  fig, axes = plt.subplots(2, 5, figsize=(25, 8))

for i, (frame, detections) in enumerate(zip(images_tensor, all_detections)):
ax = axes[i // 5, i % 5]
img = frame.squeeze(0).detach().cpu()
ax.imshow(img, aspect='auto')
ax.scatter(
detections[:, 1], detections[:, 0], marker="+", c="r", alpha=0.5
)
ax.set_xlim(0, img.shape[1])
ax.set_ylim(img.shape[0], 0) # inverted y, since image origin is top-left
ax.axis("off")
ax.set_title(f"Frame {i}")

plt.show()