Review process: option to review only some classifications

[YvanSG]

Feature
Option to review results for some classifications only

Additional Detail
On starting the review process, use a checklist to select which classifications to review: Animals are usually pretty straightforward and correct so can model can be trusted on that, and users may not want to review the images. I found issues were mostly with Multiples when only one animal was present but shadows made it look like animals; or with images classified as Empty though animals were present (small birds, reptiles). Users may want to review Persons in case they want to look at poaching, etc. Giving the option to choose would save time.

[petargyurov]

@YvanSG Just so I understand, you are requesting a filter feature to only review a portion of the images?

Example:

• Run detection...
• Results: 10 images classified as animal, 5 as person
• Filter only to show person images in review
• The 10 animal images get auto-sorted
• The 5 person images undergo human review

Did I understand correctly?

[YvanSG]

Yes, you understood correctly. The process could be:

• Run detection...
• Results: 10 images classified as animal, 5 as person
• Options for Review: list of classifications with a checkbox for each classification ([ ] animal [x] person)
• animal get auto-sorted
• person undergo review

Does that make sense?

[Jason-vdw]

Hi @petargyurov, your GUI is great!
I've just started using it for the Takahe Team looking at images to determine presence and number of predators at certain sites. I agree the above review option could be a great feature.
I have found most empty files to be fine, but often you get false positives with bait cage. Being able to review just animals to remove false positives would save time. Also, I review animal images and manually record species. If it was possible to have a drop down list where you could select most common pest species (ie possum, cat, stoat, other), and this is written to the csv? Not sure if this is possible, but it would be awesome. Thanks for your work on this!