Super.AI provides an ever-growing selection of output data types. Below you can find a list of our currently available data types, along with a brief description and an example project type for each. If you don’t see the output type that you’re interested in, reach out to us as we can probably work something out.
A super.AI project type will take a certain input data type and produce a specific output data type. Some projects take more than one input data type and/or use more than one output data type.
|Data type||Description||Example project type|
|Text||Text metadata is added to the input data||Image captioning|
|Number||Numerical metadata is added to the input data. This can be a float value, e.g., 2.3.||Image similarity|
|Integer||Numerical metadata is added to the input data. This can only be a whole number, e.g., 3.||Image counting|
|Bounding box||A 2D box is drawn over objects of interest in an image or video.||Image bounding box|
|Binary choice||Either true or false. Only one option can be selected.||Product feature verification|
|Exclusive choice||The most appropriate option is selected from a list. Only one option can be selected—just like with a radio-button list.||Image categorization|
|Multiple choice||The most appropriate options are selected from a list. Multiple options can be selected—just like with a checkbox list.||Image tagging|
|Text annotation||Metadata is added to text content. For example, labeling nouns or company names.||Named entity recognition|
|Object||A combination of different labeling methods are used to produce a single label. For example, a collection of text transcription fields.||URL transcription|
Updated 6 months ago