4

By "picture corpus" I mean the picture collection each illustrating one word in the usually used vocabulary, such as words in Google 10000 English I found on GitHub.

For instance: vocabulary
enter image description here

source: Youdao

I've crawled 2033 pictures for 2033 words (mostly nouns) in the aforementioned Google wordlist from Youdao, but it's not enough.

Improve this question
5
  • 1
    Here is the link of the Google 10000 English: github.com/first20hours/google-10000-english
    – Lerner Zhang
    Commented May 24, 2016 at 3:36
  • So you want an image for most nouns, adjectives and verbs? What about adverbs? I also noticed there are names of people and places in the Google 10000 words file. Do you want those too? Search for "jeremy" or "tanzania".
    – Bulrush
    Commented May 25, 2016 at 15:22
  • No, I need mainly nouns. For the adjectives and verbs, only if the pictures are illustrative. But hardly can they be so, I suppose.
    – Lerner Zhang
    Commented May 27, 2016 at 0:31
  • Acturally I use the words in the 20k file. Names and places are needed since they're popularly used if listed there.
    – Lerner Zhang
    Commented May 27, 2016 at 0:43
  • I found another corpus: visualgenome.org
    – Lerner Zhang
    Commented Feb 24, 2021 at 16:29

3 Answers 3

Reset to default
6

You may be able to use the International Picture Naming Project (IPNP), but I think the total pictures are way less than 10k.

enter image description here

Here's a link to the query tool.

You may get lucky searching through "image recognition" research websites. Here's one compilation of online resources, and a particular one.

Improve this answer
1
  • 1
    Thanks. Pictures from IPNP seem too simple, but what I get via the links at the end of your answer is better.
    – Lerner Zhang
    Commented May 30, 2016 at 1:53
3

Wikidata has linked various concepts and lexemes to images. For lexemes and their senses there are not that many images yet. For concepts there are many, but many are proper nouns or complex concepts, e.g., "Berlin" or "Chile at the 1956 Winter Olympics"

For Wikidata lexemes a query with the Wikidata Query Service at https://query.wikidata.org may be formulated as this:

#defaultView:ImageGrid
SELECT ?lexeme ?lemma ?image {
  ?lexeme ontolex:sense / wdt:P18 ?image ;
          wikibase:lemma ?lemma ;
          wikibase:lexicalCategory wd:Q1084 ;  # nouns
          dct:language wd:Q9035 .  # language = Danish
}

Here the language is set to Danish. The result is available here: http://tinyurl.com/yb8aeks6 You can edit the query from the link in the lower left corner.

A reformulation using sense and their links to ordinary Wikidata items (concepts) looks like this with the English language:

#defaultView:ImageGrid
SELECT ?lexeme ?lemma ?image {
  ?lexeme ontolex:sense / wdt:P5137? / wdt:P18 ?image ;
          wikibase:lemma ?lemma ;
          wikibase:lexicalCategory wd:Q1084 ;  # nouns
          dct:language wd:Q1860 .  # language = English
}

A result is available here: http://tinyurl.com/ybuuy45g

Improve this answer
0

In the VisualGenome there are 75,729 unique objects with labels, and they provide APIs for easy use and a tutorial.

To get an image word pair one only needs to crop the picture and get the corresponding region.

Improve this answer
1
  • And Wit would also be helpful.
    – Lerner Zhang
    Commented Mar 6, 2021 at 6:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.