Make it easier to get images into Ditto.
When I import a CSV of items, I'd like to be able to have a field in my CSV for the image location of each item, and for Ditto to ingest those links directly as the source of images which match the items listed.
I realize there is santization needed of remote objects which we'd download from random sources.
What would you like to see?
IMPORT -> feed a CSV file with a field called "IMAGE URL" (or perhaps some other normalized variant)
Ditto fetches individual item images and associates them with item records.
This could obviate the need to take photos/scan, if we allow the user to seed with existing image URLs.
Perhaps we could add a function to do our own image search for a match on this item which is described by the fields of the CSV used for uploading? (i.e. auto schema/category detection and mapping to item photo).
Anything you have tried or considered?
I'm just going to try to upload a CSV of items and realized that I don't have pictures of those items and would have to go through a lot of scanning to get them. Then I thought about the case where someone has already painstakingly photographed their stuff, but it'd be a pain to correlate each photo to a single item for future uploading. If we can offer a hint in the CSV for where the image is located, we should be able to ingest it and associate it with the item -- maybe also providing a new signal for image detection/recognition?
Which part of Ditto?
Scanning and capture
### What problem would this solve?
Make it easier to get images into Ditto.
When I import a CSV of items, I'd like to be able to have a field in my CSV for the image location of each item, and for Ditto to ingest those links directly as the source of images which match the items listed.
I realize there is santization needed of remote objects which we'd download from random sources.
### What would you like to see?
IMPORT -> feed a CSV file with a field called "IMAGE URL" (or perhaps some other normalized variant)
Ditto fetches individual item images and associates them with item records.
This could obviate the need to take photos/scan, if we allow the user to seed with existing image URLs.
Perhaps we could add a function to do our own image search for a match on this item which is described by the fields of the CSV used for uploading? (i.e. auto schema/category detection and mapping to item photo).
### Anything you have tried or considered?
I'm just going to try to upload a CSV of items and realized that I don't have pictures of those items and would have to go through a lot of scanning to get them. Then I thought about the case where someone has already painstakingly photographed their stuff, but it'd be a pain to correlate each photo to a single item for future uploading. If we can offer a hint in the CSV for where the image is located, we should be able to ingest it and associate it with the item -- maybe also providing a new signal for image detection/recognition?
### Which part of Ditto?
Scanning and capture
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
What problem would this solve?
Make it easier to get images into Ditto.
When I import a CSV of items, I'd like to be able to have a field in my CSV for the image location of each item, and for Ditto to ingest those links directly as the source of images which match the items listed.
I realize there is santization needed of remote objects which we'd download from random sources.
What would you like to see?
IMPORT -> feed a CSV file with a field called "IMAGE URL" (or perhaps some other normalized variant)
Ditto fetches individual item images and associates them with item records.
This could obviate the need to take photos/scan, if we allow the user to seed with existing image URLs.
Perhaps we could add a function to do our own image search for a match on this item which is described by the fields of the CSV used for uploading? (i.e. auto schema/category detection and mapping to item photo).
Anything you have tried or considered?
I'm just going to try to upload a CSV of items and realized that I don't have pictures of those items and would have to go through a lot of scanning to get them. Then I thought about the case where someone has already painstakingly photographed their stuff, but it'd be a pain to correlate each photo to a single item for future uploading. If we can offer a hint in the CSV for where the image is located, we should be able to ingest it and associate it with the item -- maybe also providing a new signal for image detection/recognition?
Which part of Ditto?
Scanning and capture