ShowTens : visualize torch tensors EASILY

ShowTens is a simple pytorch package that allows painless and flexible visualizations of image and video tensors.

<ADD VISUALIZATION VIDEO HERE>

Installation

pip install showtens or uv pip install showtens

Make sure torch is installed, as it is not automatically to allow you to choose the correct version for your system. See PyTorch installation instructions for more information.

Documentation

Documentation can be found at https://frotaur.github.io/ShowTens/

Usage

show methods work both in scripts (they open windows) and in jupyter notebooks (they display inline).

import torch
from showtens import show_image

image1 = torch.rand((3, 100, 100))  # (C,H,W) image
show_image(image1)  # Displays the image using matplotlib
image2 = torch.rand((4, 4, 3, 100, 100))  # (B1,B2,C,H,W), two batch dimensions
# Will display as a 4*4 grid, 2 pixel padding, white padding color:
show_image(image2, columns=4, padding=2, pad_value=1.0)

from showtens import save_image

save_image(tensor=image1, folder="saved_images", name="imagetensor")

from showtens import save_video

video1 = torch.rand((60, 3, 200, 200))
save_video(tensor=video1, folder="saved_videos", name="videotensor", fps=30)
video2 = torch.rand((4, 60, 3, 200, 200))  # (B,T,C,H,W), batch of videos
save_video(tensor=video2, folder="saved_videos", name="videobatch", fps=30, columns=2)  # 2*2 video grid

# Can also open the video in a new web browser window:
show_video(video2, fps=30)  # Opens a new browser window with the video