Scilab Home Page | Wiki | Bug Tracker | Forge | Mailing List Archives | Scilab Online Help | File Exchange
ATOMS : sciTorch Module details
Login with GitLab

sciTorch Module

A Module for Machine and Deep Learning via Interfacing with libTorch for Scilab
(13938 downloads for this version - 13938 downloads for all versions)
Details
Version
0.1.0
Author
Tan Chin Luh
Owner Organization
Bytecode (formerly Trity Technologies)
Maintainer
Chin Luh Tan
License
Creation Date
November 16, 2019
Source created on
Scilab 6.0.x
Binaries available on
Scilab 6.0.x:
Linux 64-bit Windows 64-bit
Install command
--> atomsInstall("sciTorch")
Description
            sciTorch module is built to allow more non-programmers (light-programmers) to be
able to use, and understand deep neural network. The reason of choosing libTorch
(pyTorch) as the engine is because it provides python/C++ parity (in the
roadmap) which make it much simple to be integrated seamlessly with Scilab.

This initial release wouls start a baby step of integration, which is importing
thr torch script module from the pyTorch into Scilab, and used it for
inferencing. Working with Scilab IPCV 4.1.2, it make the image preprocessing and
post processing much simpler to prepare to the image for the network, as well as
to present to the users.

This makes the module suitable for:
1. Lecturers who want to show the use of pre-trainined DNN/CNN
2. As the option to test the network before deploying to other platform
3. Understanding the how the correct input data important (data type, data
range, normalizing data, etc) to get good result
4. Simplifying the image processing and computer vision parts with the help of
Scilab IPCV module

We wish more experts would contributes to this module and make it a more
complete high level tools for Deep Neural Network.             
Files (4)
[114.56 kB]
Miscellaneous file

[116.14 MB]
Linux 64-bit binary for Scilab 6.0.x

[39.75 MB]
Source code archive

[106.02 MB]
Windows 64-bit binary for Scilab 6.0.x

News (0)
Comments (0)
Leave a comment
You must register and log in before leaving a comment.
Login with GitLab
Email notifications
Send me email when this toolbox has changes, new files or a new release.
You must register and log in before setting up notifications.