10.8 azcv describe

20210316 The contents of a photo can be described. Think of it answering the question what are the key elements of this photo. AI can do a good job.

For this first example of the Eiffel Tower, AI identifies a clock tower. If you puzzle enough you can see the circular arch at the bottom of the tower looking something like a clock face. We can all make incorrect observations at times. The confidence of the descriptions are in the 66% to 71% range, which is not particularly strong. The fact of it being the Eiffel Tower is at least consistent.

$ ml describe azcv https://bit.ly/3bMQX1E
0.71,a large clock tower towering over Eiffel Tower
0.66,a large clock tower towering over the city of london with Eiffel Tower in the background
0.66,the tower of the city with Eiffel Tower in the background

With even more confidences, Uluru is also described somewhat accurately, though I would not have described a canyon there. Sounds like the model may have been trained with images from the US expanse.

$ ml describe azcv https://bit.ly/3cqDonC
0.92,a canyon with a sunset in the background with Uluru in the background
0.88,a view of a canyon with a sunset in the background with Uluru in the background
0.86,a close up of a canyon with a sunset in the background with Uluru in the background


Your donation will support ongoing availability and give you access to the PDF version of this book. Desktop Survival Guides include Data Science, GNU/Linux, and MLHub. Books available on Amazon include Data Mining with Rattle and Essentials of Data Science. Popular open source software includes rattle, wajig, and mlhub. Hosted by Togaware, a pioneer of free and open source software since 1984. Copyright © 1995-2021 Graham.Williams@togaware.com Creative Commons Attribution-ShareAlike 4.0.