Is there a way to stop image warping in google colab - image

When I switched from anaconda to jupyter notebook the image inputed warped from:
To
Is there any way to solve this?

Related

Vertex AI Jupyter notebook keep setting up proxy for Jupyter Lab

When I clicked the Vertex AI and try to access the User-managed notebooks, it keep setting up the proxy for JupyterLab and not allowing me to delete, reset or stop the notebook for at least 4 hours.
I have looked up the trouble shooting doc but didn't a solution for it.

Image slider in tensorboard automatically goes to end while user scrolls through images

I am using TensorBoard 2.2.0 to visualize my input image to a neural network in PyTorch. After running the code and I try to go through the recorded image data. However, when I try to browse it the image slider automatically jumps from the current step to the last step after some time. Any reason why this could be happening.
I am launching the tensorboard using this command -
tensorboard --logdir runs --samples_per_plugin images=0
I fell through a similar dilemma. You could turn off the relead after 30s option from settings in the tensorboard. GitHub issue on the same

matplotlib animations from a remote hosted ipython notebook

My home desktop is the guts of my workstation (good CPU + memory). I often ssh in from my chromebook and port forward ipython notebooks in order to work remotely.
However, I have run into issues with displaying matplotlib animations remotely (static plots work fine). The animation appears as a static image. (The same code creates an animated plot when run on the host machine.) I know the animation is working as I can save it off into an MP4 (and the video is correct).
My current workaround is to save the video as an MP4, push it to my github page, then download the MP4 on my chromebook, and use the chrome os's native video player. This workaround sucks!
Does anybody have a solution to view animated plots in a remote hosted ipython notebook?
Or does anybody have an improvement to my current workaround?
Any thoughts are appreciated.
%matplotlib inline
import matplotlib.pyplot as plt
import numpy as np
import time
from IPython.core.display import clear_output,display
f,ax = plt.subplots(nrows=1,ncols=1)
plt.xlim(0,10)
plt.ylim(0,100)
for i in range(10):
time.sleep(0.1)
clear_output(wait=True)
ax.plot(i,i**2,'bo')
display(f)
plt.close()

Can't get Java3D going on ImageJ 3DViewer

OS: Mac OS 10.11.1
I installed ImageJ and was going to display the example image "flybrain.tif" using "3D Viewer" under Plugins -> 3D. However, the 3D viewer doesn't launch.
I searched around and realised that it could be a problem from java3d, so downloaded java3d from https://java3d.java.net/binary-builds.html. I placed the three *.jar from java3d under /Library/Java/JavaVirtualMachines/jdk1.8.0_65.jdk/Contents/Home/jre/lib
and added the paths to the ~/.bash_profile.
The 3D viewer still doesn't start. Anything else I can try? Many thanks!
I have a similar, not identical setup (Mac OS 10.11.2, jdk 1.8.0_66) and I am able to open the Fly Brain in 3D viewer using Fiji.
In Fiji, I activated the update sites for ImageJ, Fiji, and 3D.
You might also try posting the question at the ImageJ forum.
Hope this helps.
Problem solved by downloading the fiji-jogl-java3d-20151006220121.zip file from http://forum.imagej.net/t/java-3d-progress-and-next-steps/135. The 3D viewer works and I am able to record the 3D video again. Thank you every one, #TSwayne #Jan-Eglinger #gouessej

Kinect for Windows SDK vs. SimpleOpenNI RGB image quality

I'm comapring these Kinect libraries on windows:
a) Kinect for Windows SDK for Processing (http://www.magicandlove.com/blog/2012/09/05/kinect-for-processing-library-page/)
b) SimpleOpenNI (http://code.google.com/p/simple-openni/)
Link to full image: http://i.imgur.com/mghSM.jpg
Does anyone noticed the difference in RGB image coming out of these wrappers?
SimpleOpenNI is more pixelated and has more noises. Is it possible to get better image in SimpleOpenNI?
Thanks for any tips!

Resources