I tried it, it gave a poor quality image that wasn't even what I asked for. I then asked for a correction, and it gave me another faulty image. Doesn't seem to be there
Gemini 2.5 Flash Image
321–330 of 504 posts
Re: Gemini 2.5 Flash Image
#322Earlier quoted context omitted.
All of the defects you have listed can be automatically fixed by using a film scanner with ICE and a software that automatically performs the scan and the restoration like Vuescan. Feeding hundreds (thousands?) of photos to an experimental proprietary cloud AI that will give you back subpar compressed pictures with who knows how many strange artifacts seems unnecessary
I scanned everything into 48-bit RAW and treat those as the originals, including the IR scan for ICE and a lower quality scan of the metadata. The problem is sharing them - important images I manually repair and export as JPEG which is time consuming (15-30 minutes per image, there are about 14000 total) so if its "generic family gathering picture #8228" I would rather let AI repair it, assuming it doesn't butcher fa…
How did you get the 49bit and ICE data separately? Did you double scan everything?
I'm scanning my parents photos at the moment.
Re: Gemini 2.5 Flash Image
#323At $0.02 per image, it's prohibitively expensive for many use-cases. For comparison, the cheapest Flux model (Schnell) is $0.003 per image.
yes, too expensive for my use case.
Service Cost per Image Cost per 1,000 Images
Flux Schnell $0.003 $3.00
Gemini 1.5 Flash $0.039 $39.00Re: Gemini 2.5 Flash Image
#324Re: Gemini 2.5 Flash Image
#325Are men not attractive? Or perhaps for Google, this blog is a targeted content? But who is it targeting? I would like to see the reasoning behind using all women images (at the least the top/first ones) to show off the model capabilities. I have noticed this trend in the image manipulation business a lot.
Because tech is largely male dominated and has inherent sexism/patriarchy and images of women, especially conventionally attractive ones, has the perception of aiding sales. Also women are seen as more cooperative and submissive, hence so many home assistants and AI being women's voices/femme coded.
Thank you for saying that. When I posted that GP comment - it got immediately downvoted and I couldn't even see my comment on the thread. I kind of expected to get it tagged 'meta/off-topic' and removed.
The way I see it - corporations would like to exploit prejudices for revenue. Of course, this is not something new. But it is a societal issue and the corporate world is playing a large role in it.
For context this was the original link - https://deepmind.google/models/gemini/image/
Re: Gemini 2.5 Flash Image
#326Are men not attractive? Or perhaps for Google, this blog is a targeted content? But who is it targeting? I would like to see the reasoning behind using all women images (at the least the top/first ones) to show off the model capabilities. I have noticed this trend in the image manipulation business a lot.
Did they update the blog post? For me, of the 4 examples in the post, 2 are men and 2 are women.
This was the original link - https://deepmind.google/models/gemini/image/
Re: Gemini 2.5 Flash Image
#327I digitised our family photos but a lot of them were damaged (shifted colours, spills, fingerprints on film, spots) that are difficult to correct for so many images. I've been waiting for image gen to catch up enough to be able to repair them all in bulk without changing details, especially faces. This looks very good at restoring images without altering details or adding them where they are missing, so it might fina…
All of the defects you have listed can be automatically fixed by using a film scanner with ICE and a software that automatically performs the scan and the restoration like Vuescan. Feeding hundreds (thousands?) of photos to an experimental proprietary cloud AI that will give you back subpar compressed pictures with who knows how many strange artifacts seems unnecessary
Vuescan is terrible. SilverFast has better defaults. But nothing beats the orig Nikon scan software when using ICE. It does a great job of removing dust, fingerprints etc Even when you zoom in. VS what iSRD does in SilverFast, which if you zoom in and compare the 2. iSRD kinda smooches/blurs the infrared defects whereas Nikon Scan clones the surrounding parts, which usually looks very good when zooming in.
Both Silverfast and Nikon Scan methods look great when zoomed out. I never tried Vuescan's infrared option. I just felt the positive colors it produced looks wrong/"dead".
Re: Gemini 2.5 Flash Image
#328Earlier quoted context omitted.
If you compare to the amount of effort required in Photoshop to achieve the same results, still a vast improvement
Vibe coding might not be real, but vibe graphics design certainly is. https://imgur.com/a/internet-DWzJ26B Anyone can make images and video now.
Sure, now the client wants 130 edits without losing coherency with the original. What does a vibe designer do? Just keep re-prompting and re-generating until it works? Sounds hard to me.
Re: Gemini 2.5 Flash Image
#329I've had a task in mind for a while now that I've wanted to do with this latest crop of very capable instruction-following image editors. Without going into detail, basically the task boils down to, "generate exactly image 1, but replace object A with the object depicted in image 2." Where image 2 is some front-facing generic version, ideally I want the model to place this object perfectly in the scene, replacing the…
Yes! There is a model called ACE++ from Alibaba that is specifically trained to replace masked areas with a reference image. We use it in https://phind.design . It does seem like a very esoteric and uncommon task though.
Oh cool thanks I haven't come across that one, I'll give it a shot.
Re: Gemini 2.5 Flash Image
#330Earlier quoted context omitted.
> This is the gpt 4 moment for image editing models. No it's not. We've had rich editing capabilities since gpt-image-1, this is just faster and looks better than the (endearingly? called) "piss filter". Flux Kontext, SeedEdit, and Qwen Edit are all also image editing models that are robustly capable. Qwen Edit especially. Flux Kontext and Qwen are also possible to fine tune and run locally. Qwen (and its video gen s…
In other words, this is the gpt 4 moment for image editing models. Gpt4 isn't "fundamentally different" from gpt3.5. It's just better. That's the exact point the parent commenter was trying to make.
I'd say it's more like comparing Sonnet 3.5 to Sonnet 4. GPT-4 was a rather fundamental improvement. It jumped to professional applications compared to the only causal use you could use ChatGPT 3.5 for.