What Wildbook are you working in?
Amphibian and Reptile Wildbook
What is the entire URL out of the browser, exactly where the error occurred?
Can you describe what the issue is you’re experiencing?
Very high frequency of “Image Analysis has returned and no match was found.” despite high number of encounters in the project.
Can you provide steps on how to reproduce what you’re experiencing?
I have uploaded ~160 photos of adult fire salamanders to the project “Uni Bielefeld - Adults Kottenforst” using the bulk upload. I sent those pictures to identification against all locations (not just Kottenforst). When checking the matching results, I almost exclusively got: “Image Analysis has returned and no match was found.”. I find this suspicious snce the project now has 464 encounters but still there were only very few possible matches suggested. Is this correct or is there a problem with the AI or the image quality?
If this is a bulk import report, send the spreadsheet to firstname.lastname@example.org with the email subject line matching your bug report
Can you please provide a link to a match page showing this result?
This is a valid result. HotSpotter considered 2405 potential matches and ruled all of them out as unlikely matches. There was no score above 0. Unless it truly missed a match, this is HotSpotter indicating no potential matches and not wasting your time with false positives.
Hi @jason ,
thank you for your response. My colleague @Pia uploaded ~50 photos in March. In only 4 cases, she received the message “Image Analysis has returned and no match was found.” .
However, in my case, I got this message for almost every encounter out of 160. This got me wondering if there’s something wrong with either my photos (however, they do not look much different to Pia’s) or the matching AI.
Hi @jason ,
I have just tried out the “start another match” function in the encounter tab of the project " Uni Bielefeld - Adults Kottenforst" with the tip of @Pia (see: Amphibian and Reptile Wildbook | Login). Under location ID, I selected only Kottenforst (and its sub-locations). Now I have received 12 possible matches (see: Amphibian and Reptile Wildbook). This might have done trick!
Instead of testing against 2405 candidates (see for example another one: Amphibian and Reptile Wildbook), the encounter was tested against 42 candidates (Amphibian and Reptile Wildbook).
Do you think, I should upload the encounters again, this time specifying the location ID as “Kottenforst”?
If that’s where they were sighted, then adding the location ID will help focus matching on likely candidates in the area rather than all possible individuals across every study site. You can still match anything later too.
Thanks @jason , I will update this thread if I will receive more possible matches now or not.