Recent posts

[zero]

#1

AutoTagger and AI / Re: Best Strategy for 10,000 U...

Last post by Stenis - Today at 01:11:14 AM

Hello IMatch Community,

I'm a new IMatch user (and new Mini PC owner, specifically for IMatch as a Mac user!) and am incredibly impressed with its depth after just a few hours. I'm looking for guidance on tackling a large project: ~10,000 scanned family photos, currently with zero metadata.

My goal is to comprehensively tag each photo with keywords, locations, events, descriptions, and people.

I. Overall Workflow & Batch Processing Strategy:
I'm aiming for maximum efficiency. The Face Recognition tool is already proving very effective for "low-hanging fruit."

• Phased vs. Per-Photo Approach:
  
  • Would you recommend a phased approach (e.g., Pass 1: All faces; Pass 2: Locations/Events; Pass 3: Keywords/Descriptions)?
  • Or, is it more efficient to process each photo completely in one go?
• Handling Disorganized Scans: My scans are completely unsorted. Photos from the same event might be thousands of images apart.
  
  • What IMatch features or strategies (e.g., Collections, Stacks, Timeline use, specific batch processing) are best for grouping these scattered event photos as I discover them?
  • Are there particular "first pass" metadata fields (e.g., rough dates, initial event guesses) that make subsequent, more detailed tagging easier within IMatch?
• Adding "Year Taken" Information: Many photos will only have an estimated year.
  
  • What's the recommended IMatch workflow for adding an approximate year to each photo?
  • Should I, for example, create a custom tag for "Estimated Year," populate that, and then use a tool like TimeWiz to batch apply this to the appropriate EXIF/IPTC date fields? Or is there a more direct method?
• AI for Date Estimation:
  
  • Is there any existing or potential workflow (perhaps involving scripting or external tools integrated with IMatch) where AI could assist in guessing a photo's date? For instance, if IMatch knows the people in the photo and their birthdates, could AI leverage this to estimate an age for the subjects and thus a likely year for the photo?
• General Setup: Any other IMatch setup tips or considerations for a project of this scale to ensure success and avoid rework?

II. AI Autotagger (via Variables & External AI - e.g., Gemini Flash):
I've successfully used person variables to feed AI prompts for descriptions, which is amazing! However, I've encountered an issue:

• Action Misattribution: The AI (Gemini Flash) often knows who is in the photo (from the variables) but arbitrarily assigns actions to the wrong person (e.g., says "Ben is on the swing" when it was Jim, even if Ted was also present).
• Feeding Face Region Data to AI:
  
  • Is there a way within IMatch, or a known workflow, to provide the AI with face region data (i.e., bounding boxes/coordinates for each recognized person) alongside the names? This could potentially allow the AI to correctly associate actions with the specific individuals performing them.
  • Maybe I could export versions of photos with face annotations burned in, have the AI process those, and then somehow map the generated metadata back to the originals using IMatch's stacking or versioning features? Or is there a more integrated solution?
• Alternative AI Prompting/Tools: Are there other IMatch features or prompting strategies I should explore to improve the accuracy of AI-generated descriptions, especially regarding who is doing what? Or any other ideas of how to use AI to make this whole process easier?

Thanks so much for any guidance you can offer.

Why not read through some of the treads here about face recognition, prompt engineering and our joint experiences from using the different AI-models Autotagger supports and not the least how to handle keywords before you start and the Help-texts about these topics. If you don't especially keywords will get unessessarily messy.

It is not a good idea just to let Autotagger lose on its own. You will probably need to do some testing with prompting. If you haven't confirmed the identity of people on the pictures the texts will get written unessessarily unpersonal for example.

If Location data is imporant for you it will be a good timesaver to use Google Map API too with Reverse geocoding. It will automatically fill most of the location tags.

Learn how to use the metadataform efficiently.

It is almost always a good strategy to organize all your picture folders under one single topfolder and point at that at the initial indexing.

It is also common to add general static data first and then let Autotagger and the reverse geocoding do the rest and then finish  with adding specific data and correct Autotagger if you feel it us needed.

Since  most of my pictures are from travelling I always use to add a "Mandatory" first line in Descriptions that might look like this in the Ad hoc prompt - opens with F7:

Mandatory text: Paris France 2012 - and after that Autotagger writes.

There is also a very important check box in that dialogbox that let you update a single picture or let Autotagger pushvthe same text on all selected pictures.

If you don't like a common text you get it might be an idea to let Autotagger create few and pick the best and then copy with Ctrl+C and paste to the others with Shift+Ctrl+V.

#2

General Discussion and Questions / Re: Why not default to iMatch ...

Last post by Mario - June 14, 2025, 06:21:14 PM

As indicated, I won't be on Win10 much longer anyway, so I'll see how things go with Win11.

D-500 files work here just fine, with only the standard WIC codecs installed. Should work for you too.

#3

General Discussion and Questions / Re: Google Maps API restrictio...

Last post by Stenis - June 14, 2025, 05:53:43 PM

AutoTagger needs an API key. Not sure that kind of "Alias" you have created. Restricting an API key to specific APIs is always advisable and usually automatic. Google requires you to select for what you want to use an API key. Unless you enable the correct API, nothing will work. As documented in Google Maps API Key and for Gemini they activate generativelanguage.googleapis.com, I believe.

Since the API key must be accessible to IMatch, everybody that can access your computer and start IMatch can fetch your API keys. No protection against that.

In "normal" projects the API key is stored on a server, and the server accesses Google services. In this scenario you can secure the API key via the static IP address of the server o

Since the API key must be accessible to IMatch, everybody that can access your computer and start IMatch can fetch your API keys. No protection against that.

...

This cannot work if you use the API key from a desktop application on a computer which gets a dynamic IP address assigned every day or more often. Only few "private" persons have static IP addresses. But when you have one, you can use it to lock the Google API key to your computer.

Partly correct. Google offer the alternative to a static IP address restriction to be a restriction of requests from a particular browser origin. Therefore if IMatch always requests from *.phototools.com (or whatever) the key can be restricted to calls from that address. As IMatch is effectively requesting from a browser page, knowing that browser ID would allow us to add this restriction.

Unfortunately Google does not support a hard limit, e.g. if 10$ are used, block API access until next month or until I unlock. This would be the best solution.

They used to provide a hard limit with the old charging method and $200 per month free. There was a daily warning limit that didn't interfere with usage and a monthly limit which blocked requests. I have changed my workflow since the effective reduction in the free usage but will have another look.

r server farm.

This cannot work if you use the API key from a desktop application on a computer which gets a dynamic IP address assigned every day or more often. Only few "private" persons have static IP addresses. But when you have one, you can use it to lock the Google API key to your computer.

A bit of overkill, though.

Only you know your API key. You can change it when you want (and update it in IMatch accordingly). Configure Google to send emails to at least two of your email addresses when your budget is exceeded, s a warning in case you somehow "lose" your API key.

Unfortunately Google does not support a hard limit, e.g. if 10$ are used, block API access until next month or until I unlock. This would be the best solution.

Wouldn't it be difficult to spend 200 USD with Autotagger and GEMINI API?

I prefer the OpenAI solution with paying in advance upfront. Then this that happened with Deep Search on Google just can't happen. Of this reason I have ditched Google because that is an example of greed.

I have another from Greece where some has complained over a restaurant overcharging people heavily for a meal (around 400 Euro) without giving them a choice.

Hidden surprises are not nice but turning tricks like this just works once. I will just continue to use Google Map services. I woun't need Google Gemini because OpenAI  works better for me since Gemini seems to miss to update my ARW-rawfiles from time to time anyway. Bye Bye Gemini.

#4

General Discussion and Questions / Re: Why not default to iMatch ...

Last post by sybersitizen - June 14, 2025, 05:00:59 PM

As indicated, I won't be on Win10 much longer anyway, so I'll see how things go with Win11.

#5

General Discussion and Questions / Re: Why not default to iMatch ...

Last post by Mario - June 14, 2025, 04:37:40 PM

If software such as the Nikon WIC codecs is discontinued, no longer maintained, removed from the Nikon website and relocated to an Archive server, I would not consider using it. These WIC codecs operate under your user account and have the same permissions as you do. Any undetected or unfixed security issue could potentially be exploited by malicious actors.

I understand that you have files stored in a Nikon NEF variant not supported by the WIC codecs included with Windows 10, which itself is approaching end-of-support this year.

In my test library, I have about a dozen D-500 NEF files. Upon checking, all of them display thumbnails in Windows Explorer. This indicates that the WIC codecs included in Windows 11 have been updated or are now based on LibRaw, which supports many more RAW formats than the older WIC codecs provided with Windows 10. IMatch also handles these files.

If you remain on Windows 10 and prefer not to use free software like XnView, it seems your only option is to install the legacy WIC codecs from the Nikon archive.

#6

General Discussion and Questions / Re: Why not default to iMatch ...

Last post by sybersitizen - June 14, 2025, 04:07:44 PM

Again, I don't need a replacement for File Explorer, and I know about WIC codecs and third party codecs.

My real question was about the 'ton of security problems' statement made earlier. Sure, downloading something from a questionable source can be a security problem if that's the meaning. But I've been using my FastPictureViewer codecs for years, so is there some security problem in continuing to use them now or in the future? Is there a security problem in downloading and using codecs from Nikon's own servers? If so, what is the vulnerability? That's all.

#7

General Discussion and Questions / Re: Why not default to iMatch ...

Last post by Jingo - June 14, 2025, 01:40:46 PM

I cannot recommend XYPlorer enough for a File Explorer replacement.  I've been using it for decades now and the feature set is mind blowing... great for photo browsing as well with mouse down blow up, preview windows and full thumbnail display even for RAW files.

#8

Workflow / Re: An example: Where poor key...

Last post by Mario - June 14, 2025, 01:06:40 PM

Earlier many have tried but given up when realizing how much time and efforts it used to take.

Small libraries and companies with large unannotated image archives now have a hood change, using AI to describe and tag stuff. Big AI companies support fine-tuning of models for specific purposes or to add "company know-how" for the AI to learn from, even specific vocabularies and terms can be incorporated.

Another neat thing is RAG (retrieval augmented generation). If you use LM Studio you can try this out. If you drop a text document into a Chat, LM Studio feeds it into the AI and you can ask the AI questions about it...
AnythingLLM drives this even further for local AI. Fascinating times.

#9

Workflow / Re: An example: Where poor key...

Last post by Stenis - June 14, 2025, 12:19:45 PM

.... and I have really stopped to make myself irritated on how much time I have to spend writing Keywords and Descriptions or doing the job once more in my Portfolios at Fotosidan where I host them:-)

I think the new version of Autotagger and the new AI-models ought to make it a more realistic task to build decent image archives than ever since my texts now are supported in my Fotosidan-workflows. Earlier many have tried but given up when realizing how much time and efforts it used to take. 

#10

Workflow / Re: An example: Where poor key...

Last post by Mario - June 14, 2025, 08:27:55 AM

. Today it is a "non-issue".

Yes. Unless one has very specific requirements for keywords or descriptions, we can leave that boring job to AutoTagger.

And I did a lot else since July, not only working on AutoTagger. Also did PeekView, CAI, Quick Filter ribbon, the new ribbon 
and menu system, drew over 300 vector icons, re-implemented the background processing engine for even faster performance etc.