AI imager Midjourney v5 stuns with photorealistic images—and 5-fingered hands

On Wednesday, Midjourney announced model 5 of its industrial AI picture synthesis service, which may produce photorealistic photographs at a high quality stage that some AI artwork followers are calling creepy and “too perfect.” Midjourney v5 is on the market now as an alpha take a look at for purchasers who subscribe to the Midjourney service, which is on the market via Discord.

“MJ v5 at the moment feels to me like lastly getting glasses after ignoring unhealthy eyesight for just a little bit too lengthy,” mentioned Julie Wieland, a graphic designer who often shares her Midjourney creations on Twitter. “Instantly you see every little thing in 4k, it feels weirdly overwhelming but additionally superb.”

Wieland shared a few of her Midjourney v5 generations with Ars Technica (seen under in a gallery and in the principle picture above), they usually actually present a development in picture element since Midjourney first arrived in March 2022. Model 3 debuted in August, and version 4 debuted in November. Every iteration added extra element to the generated outcomes, as our experiments present:

Midjourney works equally to picture synthesizers like Stable Diffusion and DALL-E in that it generates photographs based mostly on textual content descriptions referred to as “prompts” utilizing an AI mannequin skilled on millions of works of human-made artwork. Not too long ago, Midjourney was on the coronary heart of a copyright controversy concerning a comic book ebook that used earlier variations of the service.

After experimenting with v5 for a day, Wieland famous enhancements that embody “extremely practical” pores and skin textures and facial options; extra practical or cinematic lighting; higher reflections, glares, and shadows; extra expressive angles or overviews of a scene, and “eyes which can be nearly excellent and never wonky anymore.”

And, in fact, the fingers.

Over the previous yr, the concept AI artwork turbines cannot render fingers appropriately has grow to be one thing of a cultural trope. Notably, Midjourney v5 can generate practical human fingers pretty nicely. “Palms are appropriate more often than not, with 5 fingers as a substitute of 7-10 on one hand,” mentioned Wieland.

Within the service’s Discord launch notes, Midjourney additionally famous that v5 now responds with a “a lot wider stylistic vary” than model 4, whereas additionally being extra delicate to prompting, producing much less undesirable textual content, and providing a 2x improve in picture decision.

If there is a visible draw back to the Midjourney improve for AI artwork followers, it maybe comes from photographs that may be so practical and “excellent” that the mannequin’s precision takes away a number of the thrill of repeatedly producing AI imagery to discover a appropriate end result, what one would possibly name a “slot machine impact.” Though one Twitter consumer named Philipp Lenssen noted, “You probably have a selected picture topic in thoughts, it is nonetheless a bit like lottery. However with increased profitable probabilities than v4.”