def main():
data_dir = get_ram_config()["data_dir"]
# load images
imgs = []
paths = [data_dir / "lenna.jpg", data_dir / "cat.jpg"]
for i in range(len(paths)):
img = pil_img_to_np_array(paths[i],
desired_size=[512, 512],
expand=True)
imgs.append(torch.from_numpy(img))
imgs = torch.cat(imgs).permute(0, 3, 1, 2)
# loc = torch.Tensor(2, 2).uniform_(-1, 1)
loc = torch.from_numpy(numpy.array([[0.0, 0.0], [0.0, 0.0]]))
num_patches = 5
scale = 2
patch_size = 10
ret = GlimpseSensor.Retina(
size_first_patch=patch_size,
num_patches_per_glimpse=num_patches,
scale_factor_suc=scale,
)
glimpse = ret.foveate(imgs, loc).data.numpy()
glimpse = numpy.reshape(glimpse,
[2, num_patches, 3, patch_size, patch_size])
glimpse = numpy.transpose(glimpse, [0, 1, 3, 4, 2])
merged = []
for i in range(len(glimpse)):
g = glimpse[i]
g = list(g)
g = [np_array_to_pil_img(l) for l in g]
res = reduce(pil_merge_images, list(g))
merged.append(res)
merged = [numpy.asarray(l, dtype="float32") / 255.0 for l in merged]
fig, axs = pyplot.subplots(nrows=2, ncols=1)
for i, ax in enumerate(axs.flat):
axs[i].imshow(merged[i])
axs[i].get_xaxis().set_visible(False)
axs[i].get_yaxis().set_visible(False)
pyplot.show()
Args:
i:
"""
color = "r"
co = coords[i]
for j, ax in enumerate(axs.flat):
for p in ax.patches:
p.remove()
c = co[j]
rect = matplotlib_bounding_box(c[0], c[1], size, color)
ax.add_patch(rect)
anim = animation.FuncAnimation(fig,
update_data,
frames=num_anims,
interval=500,
repeat=True)
anim.save(str(plot_dir / f"epoch_{epoch}.gif"),
writer=animation.PillowWriter(fps=30))
# anim.save( str(plot_dir / f"epoch_{epoch}.mp4"), extra_args=["-vcodec", "h264", "-pix_fmt", "yuv420p"] ) # save as mp4
pyplot.show()
if __name__ == "__main__":
config = get_ram_config()
plot_dir = list(config.plot_dir.iterdir())[-1]
print(plot_dir)
main(plot_dir)