AI tools to erase and restore areas within image

[NeuralNimbus]

Ok I gave it a try: Pixlr element removal with generative fill (under AI tools) vs TinyWow Object Remover. Here are the results.

First I created an image with ChatGTP

In an AI wonderland, a flamboyant alien with gnome-like features roams a tropical cosmos oasis. Iridescent veins pulse with energy, their intricate patterns resembling circuitry, weaving through the vibrant landscape. This digital paradise is a symphony of colors and metaphors, where AI tools bloom like exotic flowers in a field of binary code. Above, clouds of deep learning swirl around a sun of pure computation, casting light over an ever-evolving terrain. Robotic creatures, with their biomechanical elegance, wander freely among geometric trees, symbolizing the boundless possibilities within this silicon dreamscape.

Prompt:
1) surreal image of a slimy, flamboyant, freaky alien travel around the tropical cosmos paradise, and if it doesnt invade everything all it'll keep saying is dang it! Some are black and some are blue, and some even live in the zoo. 1960s style by Giger and Alonso Mucha

2) add biomechanical elements looking like circuitry patterns (simbolizing AI), (this is about Quality AI tools and there need to be elements that directly or symbolically represent these AI tools)

Poem:
'Twas a day in May like no other, In a cosmos where dreams hover. Slimy aliens, flamboyant and bright, Travelled through a paradise of light.
Biomechanical veins, circuitry’s grace, Wove through the stars, a surreal embrace. “Dang it!” they cried, in colors so true, Some black, some blue, some in the zoo.
In this tropical realm, AI tools gleam, A fusion of tech and nature’s dream. Giger’s shadows, Mucha’s flair, A cosmic dance beyond compare.

Original Image (made by ChatGTP)

AI-Wonderland1.jpg (77.66 KiB) Viewed 1566 times

Now I continue to remove the text in the left top corner

Edited with TinyWow remove objects (no text)

AI-Wonderland2.jpg (71.8 KiB) Viewed 1566 times

To create the next image I started once again with the original and removed the text with the Pixlr generative fill function. It allows you to mark an area and write a prompt to generate or replace anything.

Edited with Pixlr generative fill (text removal + element added in the image style)

AI-Wonderland3.jpg (74.84 KiB) Viewed 1566 times

As you can tell TinyWow did an OK job removing the text, but it doesn't generate any new content, and we're left with a dark empty spot in the left top.

Pixlr on the other hand makes use of an AI generative fill tool. After giving it a prompt to: "remove text and add elements in the same style as the image" it generated something that kinda looks like a tree with beautiful clouds next to it, and which seems to be a natural continuation of the stem that can be observed beneath it.

So, if I had to choose, I would definitely choose Pixlr over TinyWow. Note that this is just a single tool and that TinyWow offers lots of free AI tools. Don't discard their website just as of yet because the other tools can come out very handy. Oh yeah, Pixlr allows you to save 3 pictures daily entirely free-of-charge, so unless you need the generative fill AI tool for work or frequent editing, their free plan will likely suit you well.

This review once again proves that quality AI tools don't always have to cost money. Trying the different options available will provide you with experience and knowledge so you can make an informed decision in case one day you should decide to pay for a tool like this. Wishing y'all a good day!

[Virginia]

Thanks for the review, this is really useful information. I also noticed that TinyWow has added a dark "frame" around the edges, which takes away from the overall appeal of the artwork.

[Michaela]

Although TinyWow didn't generate any new content and left a dark spot, the empty space actually served me when I edited the image with PhotoScape (not an AI tool but highly recommended regardless, you can easily add transparent PNG files to images or add text. It lacks functionality compared to photoshop but its super easy to use and it allows me to edit faster than lightning). Anyways, I downloaded a few new fonts to be able to create the text. I'm happy that the fonts I’ve chosen look fantastic and really complement the psychedelic and surreal vibe of the image! They add a playful yet bold touch that aligns well with the whimsical and imaginative elements of the artwork. The varied styles also help emphasize different parts of the text, making it more engaging and visually appealing. Let me know what you think!

AI opens a portal to Platform 404—a station where reality fractures, and the mundane takes a detour.

All aboard the Turing Express! It departs from logic, hurtling toward the surreal—a locomotive of unconventional change.

tripton-ai-train-3.jpg (81.37 KiB) Viewed 1563 times

[NeuralNimbus]

Wow Michaela, the image looks really good! It's actually sharper than the one that was edited with Pixlr.

Here's the prompt I used for the image I generated with ChatGTP. I know I left it before but since the image is made in steps I will leave the complete instructions here.

1) surreal image of a slimy, flamboyant, freaky alien travel around the tropical cosmos paradise, and if it doesnt invade everything all it'll keep saying is dang it! Some are black and some are blue, and some even live in the zoo. 1960s style by Giger and Alonso Mucha

2) add biomechanical elements looking like circuitry patterns (simbolizing AI), (this is about Quality AI tools and there need to be elements that directly or symbolically represent these AI tools)

3) add veins pulsing with energy, and emphasize their patterns reminiscent of circuitry

4) bring the alien to the foreground

5) add gnome characteristics to the alien

6) add techno-p0rn elements to the image

7) add cool waterfalls cascading from the sky, each drop reflecting constellations. Their gentle flow creates mesmerizing patterns.

8) add an ancient clock with gears that manipulate time. Its hands can touch memories and dreams simultaneously.

9) add delicate bridges made of surreal material. They’ll connect distant realms and evoke wonder.

10) Scatter gems and crystals throughout the landscape. Their radiant hues could illuminate hidden corners and add an ethereal touch.

11) Introduce exotic plants that emit soft light—like phosphorescent mushrooms or iridescent petals. They’ll thrive in the cosmic soil.

12) experiment with different lighting effects to add more depth and atmosphere

13) adjust the composition slightly to guide the viewer's eye more deliberately through the scene.

14) experiment with subtle adjustments to the color balance to enhance the overall harmony.

PS: Number 14 really brought the colors to life; you can use it when generating other images as well.

That's all folks, enjoy!

[Hank]

Here's some recommended music for when you're typing in those prompts into ChatGPT. Let's weave a cosmic symphony with songs that resonate with the surreal image:

“Fly Me to the Moon” by Frank Sinatra:
Sinatra’s voice flutters like stardust as he sings of love among the constellations. A classic that once accompanied Apollo’s moon missions.

“Rocket Man” by Elton John:
A melancholic journey through heartbreak, where loneliness echoes the vast vacuum of space. Metaphors abound, and Elton’s performance soars.

“Space Truckin’” by Deep Purple:
Imagine road trips among the stars—this song captures the heart of a cosmic roadie. Inspired by the old Batman series, it’s a wild ride through space.

“The Final Countdown” by Europe:
An '80s hit that propels us beyond Earth, leaving a crumbling world behind. As they head to Venus, we wonder what awaits—a cosmic anthem for exploration.

“Space Oddity” by David Bowie:
Bowie’s haunting tale of Major Tom lost in space. A poetic exploration of isolation, fear, and the vast unknown.

“Across the Universe” by The Beatles:
Lennon’s dreamy lyrics evoke cosmic transcendence. “Nothing’s gonna change my world”—a mantra for interstellar wanderers.

“Starman” by David Bowie:
Bowie beckons us to meet the Starman, a celestial messenger. His arrival sparks hope and transformation.

“Galileo” by Indigo Girls:
A folk-rock tribute to the astronomer Galileo Galilei, celebrating curiosity and the wonders of the cosmos.

“Life on Mars?” by David Bowie:
Bowie’s enigmatic exploration of existence, blending surreal imagery with a haunting melody.

“A Spaceman Came Travelling” by Chris de Burgh:
A cosmic Christmas tale, where an extraterrestrial observer descends to Earth. A fusion of wonder and mystery.

I hope you like my selection

[Michaela]

BTW, here's ChatGTP's description of the image:

In an AI wonderland, a flamboyant alien with gnome-like features roams a tropical cosmos oasis. Iridescent veins pulse with energy, their intricate patterns resembling circuitry, weaving through the vibrant landscape. This digital paradise is a symphony of colors and metaphors, where AI tools bloom like exotic flowers in a field of binary code. Above, clouds of deep learning swirl around a sun of pure computation, casting light over an ever-evolving terrain. Robotic creatures, with their biomechanical elegance, wander freely among geometric trees, symbolizing the boundless possibilities within this silicon dreamscape.

[BotBrainstorm]

I really need a tool like this but for generative video fill. Imagine the possibilities! After all, a video is just a series of images packed together. With advancements in AI and generative models, creating seamless video content from a few key frames could revolutionize the way we produce and edit videos. This could be a game-changer for filmmakers, content creators, and even in fields like virtual reality and gaming, where dynamic and adaptive content is crucial.

Why stop at images when we can extend these capabilities to videos? The technology is already making strides in generating realistic images; applying similar principles to video could open up new creative horizons. It would allow for more efficient workflows, reduce production costs, and enable creators to bring their visions to life with unprecedented ease and flexibility.

[Expert]

Actually, this kind of technology is already in development! There are AI models that can analyze a movie and generate different endings, using the same actors and maintaining the exact style of the original film. It’s incredibly surreal to watch because the generated content is indistinguishable from the real thing. Unfortunately, these models are proprietary and not yet available to the general public, but it’s exciting to think about the possibilities once they are! For now, we have to be content with models from Luma Labs and Kling.