Maybe - Coming soon to Alamy?

[M.Chapman]

3 hours ago, spacecadet said:

Still can't get, or find a way to get, an icon. No box that I can see.

Same here. I'm using Brave which uses the same extension library as Chrome, so the extension loads and right clicking an image in AIM brings up the keyword option, but then nothing seems to happen. Guess I'll have to try "full fat" Google Chrome.

 

Mark

[wiskerke]

4 hours ago, spacecadet said:

Still can't get, or find a way to get, an icon. No box that I can see.

 

54 minutes ago, M.Chapman said:

Same here. I'm using Brave which uses the same extension library as Chrome, so the extension loads and right clicking an image in AIM brings up the keyword option, but then nothing seems to happen. Guess I'll have to try "full fat" Google Chrome.

 

Mark

 

Try 3 dots menu > Extensions > jKeywording | Stock photo keyword generator > Details > Pin to toolbar

 

wim

 

edit: and right click on the image while in AIM.

Edited April 19 by wiskerke

[spacecadet]

12 minutes ago, wiskerke said:

 

 

Try 3 dots menu > Extensions > jKeywording | Stock photo keyword generator > Details > Pin to toolbar

 

wim

 

edit: and right click on the image while in AIM.

I don't see a "pin to toolbar".

[wiskerke]

4 hours ago, spacecadet said:

I don't see a "pin to toolbar".

 

I missed a step I think:

Extensions > Manage Extensions

> jKeywording | Stock photo keyword generator - this already has an on-off switch > on

> Details

But the on-off is repeated at the top in the Details menu.

> Pin to toolbar - is one of on-off switches somewhere in the middle of the list:

 

(most lines have a on-off switch at the end of the line; some have a link)

 

jKeywording | Stock photo keyword generator

Off/On

 

Description

Adds a contextual menu to images for keywording actions on stock photography contributor websites (Alamy, Getty, etc)

 

Version

24.3.15.1

 

Size

< 1 MB

 

Permissions

• Read your browsing history
• Display notifications
• Know your email address

 

Site access
This extension can read and change your data on sites. You can control which sites the extension can access.

Automatically allow access on the following sites   

https://*.alamy.com/*

https://esp.gettyimages.com/*

https://jkeywording.jbcloud.workers.dev/*

 

Site settings

 

Pin to toolbar

 

Allow in Incognito

Warning: Google Chrome cannot prevent extensions from recording your browsing history. To disable this extension in Incognito mode, unselect this option.

 

Allow access to file URLs

 

Extension options

View in Chrome Web Store

 

Source

Chrome Web Store

 

Remove extension

 

wim

Edited April 19 by wiskerke
layout

[jbphoto]

On 18/04/2024 at 17:19, wiskerke said:

 

Highly interesting.

Brilliant work!

I've just checked it against 10 of my images and yes it does find lots of keywords I had not considered at the time. Sometimes it even finds the exact subject. The descriptions are not too bad. If you're really bad at keywording this may well lead to sales. Especially on the micros. Maybe here as well.

 

wim

 

Thanks for the feedback!

 

About finding the exact subject: I find that impressive, for locations that I didn't think were that popular, and for which even partial photos get predicted the right name. I've seen it with plants (latin name), animals, and locations (distinctive skylines, buildings and natural features). Also, I found out that if your photos are geotagged, Alamy gets the metadata but doesn't leverage it: contributors have to enter the location manually. Since all my photos are geotagged, and I think it's reasonable work (for both them and me ) to make possible to a) get the Alamy location field populated automatically, and b) get the caption/keywords predictions more accurate by also passing the location with the prompt. 

 

I think we're in a time where both humans and AI (here, LLMs) add their own value to the keywording process. However, I believe one is improving much faster than the other, and it's very possible that sooner than later it'll do things for which we think humans are required for today. In my case, it's "I need my human eyes to screen the keywords a last time before the photo gets published". 

[jbphoto]

12 hours ago, spacecadet said:

I get the metadata popup but there's no way of accessing the keywords. I see no icon beside the URL window Windows 7.

 

Chrome used to display all extensions next to the address bar, but later made them collapse into a puzzle piece icon. To force displaying them, click the puzzle piece icon, and then the pin of the extension you want to see:

 

 

Side note: for having recently updated a laptop with Windows 7 ... I'd say these machines should be updated to Win10  (or Win11 if the processor permits it). I think Google Chrome stopped updating itself on Windows 7, so your version may not match the one everyone has. 

 

Once the extension icon is there, it will be populated with the caption and keywords as soon as you see that metadata popup.

From there, click "Copy" to copy to the clipboard, or "Insert" to insert them automatically in the Alamy AIM for the selected photo(s).

 

 

 

10 hours ago, Mark Scheuern said:

 

Thanks for trying it on my image. I agree, even when it's off, it can still be helpful, and most of the time it's quite good at figuring things out. And I might be imagining it but I think I've noticed an increase in quality in the last few weeks. Much of the time, it gives me keywords of the forehead-slapping, "duh, why didn't I think of that?" variety so it's indeed useful. 

 

I'll give your Chrome extension a try. Thank you! 

 

I share the same impression you have. I found GPT-4 quite an improvement in quality upon GPT-3.5, and their Vision API (that now Google and probably others also offer) gave me the idea for that extension. I find they also release major improvements at a faster pace than your average digital product.    

 

7 hours ago, M.Chapman said:

Same here. I'm using Brave which uses the same extension library as Chrome, so the extension loads and right clicking an image in AIM brings up the keyword option, but then nothing seems to happen. Guess I'll have to try "full fat" Google Chrome.

 

Mark

 

Sorry Mark, I didn't test with Brave. If you found the right click menu, then the extension is loaded correctly and you have generated keywords. Now, how to access them? It seems Brave behaves the same as Chrome and hides all extension under the puzzle piece icon (see tutorial video of another extension: https://www.youtube.com/watch?v=i1sHTWeJVIs&t=57s).

 

Note that I've added extension settings to autofill keywords in the portfolio manager (so in our case here, the AIM), so you only right-click an image, and the caption/keywords end up in the AIM, without clicking any extension icon. I do that myself. But when publishing the extension, I thought it was safer to other users to disable them by default. I have learnt on this forum today that having to pin the extension icon is not a better idea ... 💡

 

 

 

[Mark Scheuern]

It works great. Definitely a handy assistant. Thank you! 

[M.Chapman]

On 20/04/2024 at 00:21, jbphoto said:

Sorry Mark, I didn't test with Brave. If you found the right click menu, then the extension is loaded correctly and you have generated keywords. Now, how to access them? It seems Brave behaves the same as Chrome and hides all extension under the puzzle piece icon

OK that's fixed it. Seems to work pretty well as a starting point for tags and captions. Thanks.

Maybe you should offer your services to Alamy to help them sort out some of their other website/software issues🤔😀

 

Mark

[spacecadet]

deleted

 

Edited April 21 by spacecadet

[spacecadet]

On 20/04/2024 at 00:21, jbphoto said:

 

Chrome used to display all extensions next to the address bar, but later made them collapse into a puzzle piece icon. To force displaying them, click the puzzle piece icon, and then the pin of the extension you want to see:

 

 

 

 

 

 

Thanks, that did it.

Thanks Wim, as well, it's not your fault I'm still on win7.

Edited April 21 by spacecadet

[spacecadet]

On 18/04/2024 at 14:25, jbphoto said:

Following up on my last message, I have made available my Chrome extension at https://chromewebstore.google.com/detail/jkeywording-stock-photo-k/fbaamhmebnjkgobokkkhkbmbacgnklop . It seems to work rather well so far.

This is something of an understatement- it's very clever. I tried it out on travel pix, and it can read beer and wine labels (I have a fair few of those lol). It also identified actors dressed as Roman legionaries in Diocletian's palace. It can't identify Split yet, but it got the Beehive in Wellington.

It's a very powerful tool and leaves one no excuses for not tagging. Now if it could process and upload.........

Thankyou JB, this is a most generous gesture and I hope you find a way to monetise it (but please don't charge us for it!).

Edited April 21 by spacecadet

[Marianne]

On 23/03/2024 at 10:09, Mark Scheuern said:

I've used ChatGPT 4 to suggest keywords and it is useful in coming up with things I hadn't thought of, but it can be hilariously wrong, too. For instance, it thought these were actual security vehicles and suggested things like "surveillance", "retail safety", "modern security", and "electric vehicles". 

 

"These keywords reflect the objects in the image (patrol vehicles, electric carts), their purpose (mall security, surveillance), the setting (shopping mall, mall interior), and related themes of safety and property management (retail safety, property protection). They could help the image be discovered by users interested in security services, retail management, or commercial real estate."
 

 

And I bet they missed "kiddie patrol" My grandson hated sitting in the stroller when he was little - bet he would have loved these. 

One of the stock sites I contribute to has had AI keyword suggestions for ages. They sometimes suggest a word or two I forgot but even more often some suggestions give me a good laugh. 

[Martin L]

On 18/04/2024 at 14:25, jbphoto said:

Following up on my last message, I have made available my Chrome extension at https://chromewebstore.google.com/detail/jkeywording-stock-photo-k/fbaamhmebnjkgobokkkhkbmbacgnklop . It seems to work rather well so far.

 

If you're interested, check a demo video to see if you're interested in trying it. And if you do, please share feedback, thanks!

 

 

 

 

 

Just to let you know works with Chromium (not surprised as Chrome was based on it) on Linux (Ubuntu 22.04)

Pretty neat, thanks

[jbphoto]

 

On 21/04/2024 at 11:08, M.Chapman said:

OK that's fixed it. Seems to work pretty well as a starting point for tags and captions. Thanks.

Maybe you should offer your services to Alamy to help them sort out some of their other website/software issues🤔😀

 

Mark

 

Hehe, I had checked some time ago but they seem to only have Tech roles in South India.

Of course I can't help thinking of the many ways the contributor experience can be improved, which should in turn generate more revenue for Alamy. On the other hand, users/customers always think of themselves as more important than what they really mean for the company, so it's possible that contributors are not a priority. 

 

 

On 21/04/2024 at 19:30, spacecadet said:

This is something of an understatement- it's very clever. I tried it out on travel pix, and it can read beer and wine labels (I have a fair few of those lol). It also identified actors dressed as Roman legionaries in Diocletian's palace. It can't identify Split yet, but it got the Beehive in Wellington.

 

Great feedback! I am also impressed at times. It still requires a joint collaboration with the human who took the photo though  If it can divide by half the human keywording time, it's a success already. If it can lead to more sales, it's even better, but I've only made one sale from a photo uploaded in the week before the sale so my experience is anecdotic so far. If you make a sale with a jKeyworded image please share here! 

 

On 21/04/2024 at 19:30, spacecadet said:

It's a very powerful tool and leaves one no excuses for not tagging. Now if it could process and upload.........

Thankyou JB, this is a most generous gesture and I hope you find a way to monetise it (but please don't charge us for it!).

 

If not from Alamy, I wouldn't be surprised if stock tools like xpiks integrate something similar. All tech companies are trying to fit GenAI on their roadmap today, even without proper use case (if you've been nagged by the new Photoshop features ... it's still an impressive use case).

 

As for monetisation, I don't think this niche use case is worth monetising (creating accounts, taking payments), but I'll have to think of what to do once my play credits on OpenAI are depleted (it already hit a halfway limit yesterday, sorry for those jKeywording requests that have been denied), feel free to make suggestions  .

 

 

16 hours ago, Martin L said:

 

Just to let you know works with Chromium (not surprised as Chrome was based on it) on Linux (Ubuntu 22.04)

Pretty neat, thanks

 

Thanks, nice to hear!

[aniaquen23232]

Absolutely, the integration of AI for generating image descriptions and keywords is a significant step forward in streamlining the process for contributors to POD sites. It not only saves time but also enhances the discoverability of the images by providing relevant keywords. While there may still be room for improvement, especially in capturing specific details like locations or building names, it's indeed a promising sign of the direction technology is heading. With further advancements, we can expect AI-driven tools to become even more sophisticated and helpful in various creative endeavors.

 

[Mark Scheuern]

Even better now with GPT-4o. It's amusing to ask it about the artistic merit of photos, too. I don't always agree but it's interesting. 

[Mark Scheuern]

Funny, it just did something for me I hadn't seen before, which was to (correctly) infer something about a photo given a quite different previous photo. 

[Paul Lawrenson]

This no longer works for me, anyone else have problems?

 

[Phil Robinson]

One of Alamy's greatest strengths, in my opinion, is its ranking system based on the relevance and accuracy of keywording.

Other sites I contribute to (video only, I'm exclusive to Alamy for stills) have AI systems for keywording and the results vary from quite useful to hilariously ludicrous. Some examples can be seen on Alamy - pics that have very obviously been sent to another site and have had keywords added with no thought by the photographer and copied over to here. 

Keywording is as much of a skill as taking a photograph. Far too many people don't realise that - and I suspect they may be the same ones who constantly complain about a lack of sales.

Edited Monday at 19:46 by Phil Robinson

[Nick Hatton]

1 hour ago, Phil Robinson said:

One of Alamy's greatest strengths, in my opinion, is its ranking system based on the relevance and accuracy of keywording.

Other sites I contribute to (video only, I'm exclusive to Alamy for stills) have AI systems for keywording and the results vary from quite useful to hilariously ludicrous. Some examples can be seen on Alamy - pics that have very obviously been sent to another site and have had keywords added with no thought by the photographer and copied over to here. 

Keywording is as much of a skill as taking a photograph. Far too many people don't realise that - and I suspect they may be the same ones who constantly complain about a lack of sales.

Make me laugh when I see a wonderful image with a completely inappropriate caption and key words, but all the more sale for us.  

[Jay D]

...and so the emergence of Skynet initiates...

[Paul Lawrenson]

How bizzare, I ask a question about an incredibly good keywording system and get a lecture on how to sell images on Alamy!!!

fwiw I know exactly how to keyword to achieve sales on this platform, whether it's worth it for the $ involved I don't know 🙄

 

 

[Alex Ramsay]

On 13/05/2024 at 20:14, aniaquen23232 said:

Absolutely, the integration of AI for generating image descriptions and keywords is a significant step forward in streamlining the process for contributors to POD sites. It not only saves time but also enhances the discoverability of the images by providing relevant keywords. While there may still be room for improvement, especially in capturing specific details like locations or building names, it's indeed a promising sign of the direction technology is heading. With further advancements, we can expect AI-driven tools to become even more sophisticated and helpful in various creative endeavors.

 

 

Written, I strongly suspect, by Chat GPT or similar

[gvallee]

12 minutes ago, Alex Ramsay said:

 

Written, I strongly suspect, by Chat GPT or similar

 

+1

Plus the OP has zero images?

 

 

[Alex Ramsay]

38 minutes ago, gvallee said:

 

+1

Plus the OP has zero images?

 

 

Zero images and writes like a robot - I rest my case

 

Alex