|
Compucon Machine Learning Sampler |
|
|
July 2019 |
This Compucon Machine Learning Sampler #1 is provided free of charge and obligations. It represents the development effort spent by Compucon and does not claim any originality or patent right at this stage. Please do not replicate or distribute this software folder without prior consent from Compucon.
This sampler software contains a machine learning model and 100 images. The software has been previously trained on
over 100,000 images (pictures or photos) so that it recognises many objects and is able to describe what an image shows.
The description is called the caption of an image. Each execution of this software takes seconds, and one of the stored
images will be displayed on screen as well as a caption that the model has worked out instantly. A total of 10 such
images (randomly selected from the 100 images stored) and their corresponding captions will be given. If this software
is executed again by clicking, it will produce another 10 images with captions but these images may repeat some of the
images previously displayed.
This is only a very simplistic sampler. The Compucon team will release Sampler #2 and so on for increasing complexity
over time. Enjoy!
Sample images below, Original Image(left) -> Processed Image(right):
|
