Internet service "The Wolfram Language Image Identification Project" that identifies what is shown in the picture and displays the name and related information

There are times when you want to identify what is in the picture and you do not know what it is. A convenient net service that identifies what is shown in the photo and displays the name at that time is a mathematical expression processing systemMathematicaDevelop and sellWolfram ResearchCreated by "The Wolfram Language Image Identification Project"is.

The Wolfram Language Image Identification Project
https://www.imageidentify.com/

The site has been developed for more than 25 years agoWolfram (Wolfram) languageIt is a site created with the program language called "Wolfram language" that was created for the purpose of getting to know one of the 5000 functions incorporated in the Wolfram language. Wolfram language is available in various environments such as desktop · mobile · cloud,MathematicaIt is also used as an interface language.


How to use the site is easy,Top pageDrag and drop the image you want to identify into the red frame or directly click inside the red frame to select the image. It seems that the image is mainly processed by the function called ImageIdentify in the Wolfram language.


In addition, sample images are prepared on the right side of the site without preparing images, so you can use it.


First of all, I will try to identify Tokyo Tower with this picture.

ByRen Kuo

Although the result was identified as "tower" (tower), it was not specified as Tokyo Tower, but it is a good place until the correct answer. The choice that can evaluate the accuracy of the result under "tower" "Great" (wonderful) "Could be better" (better) "Missed the point" (What is it like "what the heck?" (What's going on? ?!) Is prepared.


To evaluate the accuracy of the result as "Could be better" (better), click "Could be better". Then, the candidate "construction" (building) "artifact" (artificial object) "Eiffel Tower" (Eiffel Tower) and the input form are displayed so that "Tokyo Tower" (Tokyo Tower) Please enter accurate information and click "Submit". Candidates such as the Eiffel Tower displayed for specifying the photograph were selected from the data accumulated by the Wolfram language, and if there is a reply specifying the photograph, a mechanism to enhance the accuracy to further accumulate and specify the information It is getting.


To specify another photo, click "Try another image" in the lower right.


Next I will try using this picture with a lot of strawberries on the dish.

BySancho McCann

The result was "fruit" (fruit). Even if there are many objects to be identified, there seems to be some degree of accuracy.


Next, I will try using the pictures of corgi dogs.

ByMei

The result of this time is not a dog but "Pembroke Welsh Corgi" (Welsh Corgi Pembroke) And up to the breed. It looks similar to the corgi dog in the photoWelsh CorgicardiganIt is possible, but it is almost certain. In addition, data on Welsh · Corgi · Pembroke was additionally displayed at the bottom.


Data on Welsh Corgi - Pembroke,Question Answering SystemOnline serviceWolframAlpha"Female weight" (female weight) "female weight" (female weight) "size" (size) "life span" (average life span) "litter size" (number of babies) "hair Length "(hair length)" alternate names "(aka" history "(history)" description "(explanation)" temperament "(personality) was displayed.


Opening "history" (history) shows the history of Welsh · Corgi · Pembroke generated by Wolfram Alpha.


Finally one of Apple's founderSteve JobsI tried what I could identify with the photos. The first answer was identified as "Person" (person), then the celebrity element and picture were analyzed together and correctly identified as "Steve Jobs" (Steve Jobs). As the accuracy of specifying photos seems to change considerably depending on the photos you choose, it seems that choosing easy-to-understand photos leads to accurate judgment.


Wolfram language for knowledge-based programming