Jump to content

k1w

Members
  • Content Count

    24
  • Joined

  • Last visited

  • Days Won

    8

Everything posted by k1w

  1. Thanks! Ran the statement and 730 rows were updated. Trying to regenerate AI labels now, and will report back.
  2. I posted a question at the end of this topic, but since it didn't get an answer, and it is different enough to warrant a separate post, I am doing so here. 🙂 I am still have issues with some photos with some AI tags not able to be updated. I think that WilfriedB's SQL query from the previous thread is likely the culprit. select * from cloudvision where value = '[ ]' What I'm not sure of is the best way to manually update the rows in that table for Daminion to recognize that they need to be regenerated? Should I set column 'value' to NULL, or delete the problematic rows
  3. I am still have issues with some photos with no AI tags not able to be updated. I think that WilfriedB's SQL query with cloudvision.value = '[ ]' is likely the culprit. What I'm not sure of is the best way to manually update the rows in that table for Daminion to recognize that they need to be regenerated? Should I set column 'value' to NULL, or delete the problematic rows from table cloudvision entirely? And if I do that, are there other tables that need to be updated to keep everything in sync, or is just updated cloudvision enough? (FYI, I am quite comfortable with writing SQL)
  4. I get this error any time I select photos with no AI labels to get labels, and choose the option for "All" rather than the 5000 selected in the current window. At the time of this error it was about 21,000 photos I was attempting. I can add 5000 at a time with no problem, and can select 5000 more back to back several times and they queue properly. But if I try to choose the most Daminion prompts me for, it crashes. Is it limited to 16384 max, as suggested by the "maximum array length quota" in the error screen shot? See the end of this message for details on invoking just-i
  5. k1w

    PostgreSQL 12

    Is there any timeline for PostgreSQL 12 integration? I just tried reinstalling Daminion Server with PostgreSQL 12 available on my system, and it didn't seem to detect it, nor offer any place to manually configure.
  6. Thanks WilfriedB. I use SQL in my day job, so trying your suggestions out shouldn't be a problem. Much appreciated!
  7. Is there any way to monitor the progress of generating AI labels? I'm attempting to generate labels for a lot of unlabeled photos (now that the "search for unassigned" bug is fixed), but it seems like a bunch of photos are not getting labeled, even after several hours. It's difficult to tell whether they aren't getting queued, whether the queue is stalled, or whether they were not processed successfully. When I look at the cvis.txt log file, I see a lot of entries like: DEBUG 2020-03-19 13:03:35,751 CVis.OrchestratorActor [29] - NextTaskMessage INFO 2020-03-19 13:03:35,760 CVis.Imag
  8. Appears to be working for me as well. Thank you!
  9. Thanks, rene. My hope was to find something that would detect them automatically (or at least a high percentage of them), without having to manually go through all 2000 images (since they are scattered throughout). Plan B would be to do as you say.
  10. Hi everyone, I've discovered that with a recent import of about 2000 scanned family photos, that a small percentage of them were upside down during scanning. Rather than go through them manually, I was hoping there was some way to automatically detect and mark photos are likely to be flipped. Does Daminion (or some other software utility) have some capacity to suggest photos that are upside down? Note that I am not asking about the EXIF orientation flag, but rather detecting based on the contents of the image itself that it is wrong, perhaps using some neural net/deep learning model.
  11. Wonderful to hear, Daria! Thanks for the update. Will it handle being able to search for photos with empty tags, and thus also adding tags only to those that have not yet been processed? If you delete all tags from a photo, does that get considered as "empty" (untagged) and thus would be requested again if you attempt to generate tags for unprocessed files, or will it be treated as "no tags" (tags were generated, discarded as unwanted, so not processed again)? Or will the user have the ability to specify either state per photo? Looking forward to trying it out!
  12. Bumping this to see if there's been any progress made the issues with AI labels -- searching for empty, recovering deleted AI labels, etc.?
  13. Any update following the remote session with Wilfried?
  14. No change after completing optimization. Searching for "AI Labels is empty" matches 48449 out of 48449 items. Searching for "AI Labels is not empty" matches 16095 out of 48449 items.
  15. That is correct. 48449 total files of which 16k of them have AI labels (done by repeating batches of 1000 files at a time). I will try the optimization now and report back.
  16. Tried with the new 6.3 build that was announced today. Still seems to be a problem searching for empty AI labels:
  17. If "Unknown" is the same as Unassigned, then all 48k+ photos are listed as Unknown, even though 16k+ of them have AI Tags (as correctly detected by the "is not empty" search.
  18. Build 2030 still has problems with searching for "AI Tags" "is empty". It returns all files as matches. But it looks like "AI Tags" "is not empty" correctly finds files with AI tags.
  19. Is there any update on being able to search for only photos with no AI Tags?
  20. I was selecting a batch of 1000 files, generating AI tags, and then walking away, so I don't know exactly how long it took to complete. But on my Google console, ti's showing a peak over the past 7 days of just over 200, comparable current usage, and so overall looks a lot similar to what you posted.
  21. I just came to the forums to report the exact same issue. I've used the new build to successfully generate AI tags for a few thousand photos, but advanced search to find empty labels shows all items. Furthermore, under the Tags pane, the number of AI Tags listed as "Unknown" never goes down, even as photos are successfully tagged and appear elsewhere in the AI tags list.
  22. k1w

    AI Tagging

    With some help during a remote session tech support today, things seem much better in getting AI Tags to populate. That said, I am still seeing errors in the logs that suggest some kind of duplicate key issue. Maybe this only shows up if you try to update AI tags for a photo that was done previously? Or is this a spurious error that I should ignore? 2019-11-22 11:44:32,209 [47] ERROR PicaJet.Daminion.Service.Utils.Command.SetToProcessLabelsCommand [(null)] - Npgsql.NpgsqlException: duplicate key value violates unique constraint "cloudvision_pkey" Severity: ERROR Code: 23505 at
  23. k1w

    AI Tagging

    I am having similar problems. I have enabled AI tagging and restarted the server, turned on AI tags in the client interface, selected several photos and did Actions -> Create AI labels. I have Google Cloud Vision activated, and can see a few requests showing up in the Google console. I have also successfully generated tags for a test photo using a Python script outside of Daminion, so I know that it is capable of working. After waiting a long time, with numerous tag refreshes, no tags appear under AI tags. This occurs even with a brand new catalog I specifically created to test AI tagg
×
×
  • Create New...