Digital image annotation and retrieval systems and methods
In a digital image annotation and retrieval system, a machine learning model identifies an image feature in an image and generates a plurality of question prompts for the feature. For a particular feature, a feature annotation is generated, which can include capturing a narrative, determining a plurality of narrative units, and mapping a particular narrative unit to the identified image feature. An enriched image is generated using the generated feature annotation. The enriched image includes searchable metadata comprising the feature annotation and the plurality of question prompts.
1. At least one computer-readable medium having computer-executable instructions stored thereon, the computer-executable instructions structured, when executed by at least one processor of a computing system, to cause the computing system to perform operations comprising:
identifying an image feature in an image included in a set of images;
generating a feature annotation, comprising:
capturing a narrative, via one or more information capturing devices, the narrative comprising at least one of a video file, an audio file, or text;
based on the captured narrative, generating a transcript;
using the transcript to determine set a set of narrative units; and
mapping a particular narrative unit from the set of narrative units to the identified image feature to generate the feature annotation; and
using the generated feature annotation and based on the image, generating an enriched image,
wherein the enriched image comprises rich metadata, the rich metadata comprising the feature annotation.
2. The at least one computer-readable medium of claim 1 , causing the computing system to perform operating further comprising:
generating a feature animation by generating an image sequence comprising the enriched image,
wherein the identified image feature is visually emphasized based on the feature annotation.
3. The at least one computer-readable medium of claim 2 , causing the computing system to perform operating further comprising rendering the feature animation via a user interface while capturing the narrative.
4. The at least one computer-readable medium of claim 1 , wherein capturing the narrative is performed prior to or after identifying the image feature.
5. The at least one computer-readable medium of claim 1 , causing the computing system to perform operating further comprising:
based on an item associated with the narrative, the item comprising at least one of a particular term, tone of voice, pitch, or facial expression category, generating a sentiment indication; and
including the sentiment indication in the feature annotation.
6. The at least one computer-readable medium of claim 1 , causing the computing system to perform operating further comprising cross-linking the narrative to a plurality of images comprising the image.
7. The at least one computer-readable medium of claim 1 , causing the computing system to perform operating further comprising cross-linking the narrative to a second identified image feature within the image.
8. The at least one computer-readable medium of claim 1 , wherein the rich metadata further comprises at least one of a label associated with the identified image feature, a tag associated with the identified image feature, a time code, or geospatial metadata.
9. The at least one computer-readable medium of claim 1 , causing the computing system to perform operating further comprising using the generated feature annotation, generating, by a third machine learning model, a query.
10. The at least one computer-readable medium of claim 9 , causing the computing system to perform operating further comprising:
generating a user interface and presenting, via the user interface, on a user device, the query;
capturing a user response to the query;
generating a series of queries based on at least one of an item within the query or the captured user response; and
presenting, via the user interface, on the user device, the series of queries.
11. The at least one computer-readable medium of claim 10 , causing the computing system to perform operating further comprising:
based on the captured user response, searching a plurality of rich metadata items associated with a set of images,
wherein the set of images is automatically determined by determining one of a family tree or family group associated with at least one of a user and the identified image feature.
12. The at least one computer-readable medium of claim 11 , causing the computing system to perform operating further comprising based on the captured user response, updating the rich metadata.
13. The at least one computer-readable medium of claim 1 , causing the computing system to perform operating further comprising generating a synthetic annotation for the enriched image by combining a plurality of rich metadata features for a plurality of identified features.
14. The at least one computer-readable medium of claim 13 , causing the computing system to perform operating further comprising supplementing the synthetic annotation with additional user input.
15. The at least one computer-readable medium of claim 13 , causing the computing system to perform operating further comprising:
generating the synthetic annotation based on an additional plurality of rich metadata features generated for a different enriched image; and
associated the synthetic annotation with the enriched image and the different enriched image.
16. The at least one computer-readable medium of claim 15 , causing the computing system to perform operating further comprising determining the different enriched image by providing the identified image feature to a fourth machine learning model trained to identify related images based on an item in the identified image feature.