Writing a good image of artificial intelligence is not difficult, but you need these basic elements

One of the first things you learned during the test Image generators Amnesty International It is that there are a lot of things that can get worse when you try to get the image you see in your head to appear on your screen. If you have used an AI image or a video generator, you know what I mean.

I spent last year in the test and review of various Amnesty International photos, and I created hundreds of photos through services such as Google banana modeland Midjourney and D.. But the pictures that I created were not all winners. It was a handful of them Frankly terrifying. But all of my tests forced me to know that the best way to avoid creating an artificial artificial intelligence is to use a good claim.

Immediate engineering, as experts call it, know the words to be used to get AI products to do what you want. For artificial intelligence images, this means creating a total description of what you want, to what is beyond letters and preparation. Regardless of the service you use, there are basic elements you need in each directed to get the best results. This is especially important if the generator does not have a lot of editing tools, such as the ability to download reference images or repair of exotic hallucinations.

Writing a good AI image is not difficult, but it may take a little more work than you expect. These are the best expert advice to formulate the correct claim, including some useful phrases for using them and common mistakes to avoid them.

Start with these three elements

When you write your face for the first time, you may feel tired or you are not sure where you start. I was there, the best place to start is with the basics. These are the three elements necessary for all fast needs. Once you have something for each of these, you can build it from there.

• Verses and elements in the scene
• Preparation or where happens
• Dimensions, such as image, landscape, or specific percentage (3: 2, 16: 9, etc.)

You may tend to add some exclusion properties in a router, or things you don’t want in your image. I would like to be against him. Even the most exciting generator is likely to ignore it, or worse than that, it offends the claim and includes something specifically you requested. If you want to eliminate an element from one image, it is usually easier to do this in the editing phase rather than the original claim.

Select the style panel and colors you want

Besides “who, what and where” in the primary router, you will need to direct the generator towards a specific style. Here are some of the most popular patterns of artificial intelligence images.

• Realism: the closest to real life. Amnesty International’s photo generators are not great in this, but it is worth trying.
• Photo gallery: Like real, but brighter and brighter.
• Product features: Individual items emphasize the background or scene.
• Cartoon: fun, bright and usually less detailed.
• Clarification: Sewers of paintings, pencil drawings.
• Usernaging games/game: more advanced than animation, and sometimes resemble animation.

Include specific colors you want as well. If you are not difficult to satisfy the fine shadows you want, you can still drive the generator below the right path by determining whether you want warm or cold tones.

You will need different patterns for different projects. Realistic artificial intelligence images are likely to be more suitable for professional environments than images similar to animation, but they may not be suitable for creative inspiration. The illustrations may be better for more projects directed towards detail, such as building brainstorming ideas, and good games for the first repetition of new characters and worlds.

The aesthetic, vitality and passion class

Take a step forward and include a description of comprehensive aesthetic or fiber. This can help upload your photos and access this additional layer of details. These details are a starting point to make you on the field for what you want without overwhelming the birth with a new demand. Here are some common options that must be included in your demands.

• a summary
• Anime
• middle ages
• Reactionary
• anesthetic
• Glow, Neon
• geometric
• Drawing, brush kick, oily paint
• comic
• Nuer
• classic
• Impressionism
• Simple, minimal
• Imagination, science fiction
• high tech
• surrealist

If none of these aesthetics feel right, try choosing the closest and construction from there. Includes textures, time and monuments. If you care less about the specified style but you want to ensure a specific emotional response, try describing it. The emotional temperature of the scene can often be described as the generator towards a specific type of visual appearance. For example, happy scenes tend to have bright colors and a warm feeling, regardless of whether the incident or illustrations. The stressful scenes may have more details, cold tones and a feeling that the generator may show you better with a fictional or unrealistic aesthetic.

You can try to use more specific aesthetics or pop culture, but there is no guarantee that the generator will understand and adhere to. For example, you may want to think about translating “Cottage Core Coastal Grandmother” into “Style Vintage Style with light, drawn, feeling using pastel blues and neutral colors”. He gets the same idea with more specific instructions.

My artificial intelligence is still incorrect. What now?

Even with a well -written router, artificial intelligence generators are not perfect and will get some factors. Text technology is progressing to the image, but it is still under progress.

The change in your claim is the fastest way to explore big problems. But if the problems persist, try to narrow the exact error in the pictures and follow the problem to the place from which it may come. For example, if your photos are not sufficiently professional to present them, it may be because the style or aesthetic included in your demands is incorrect. Even making smaller changes on your pre -settings, such as the dimensions of the image, can make a big difference in the final results.

Many AI’s photo services offer post -generation editing tools that can help you fix smaller errors. Services are more directed towards professional creators such as Adobe firefly You have wide tools. More beginners -friendly programs run maneuvering, with Leonardo More than others, then Midjourney With an average amount, with Canva Busty.

However, it may be frustrated not to get what you want after a lot of work. The most frustrated is that the best thing to do is start again. Reset your settings to virtual, rethink claims and start again you can feel the desire to go back. But when nothing else does, the last resort can be good.

At the end of the day, artificial intelligence generators are not alternatives to creators. They are like other photo editing programs: you need to spend time learn about your program, and understand how it works and editing abilities. Once you deal with your program, you will have a good understanding of the type of claims that make the best results. These tips will help you get close to what you want.

For more, check out Best Chatbots Amnesty International and What do you know about video generators of artificial intelligence.