Photo credit: Julie Wieland
Before Midjourney AI v5, most artificial intelligence tools just weren’t able to generate photorealistic human hands due to their complicated geometry and their lack of universal collection of lines or shapes it can use to identify the shape. This version was trained using Google Cloud’s v4 TPUs with Deepmind’s JAX, as well as inference running on large GPU clusters.
Since the development of Midjourney V5 is still ongoing, the team is collecting human feedback by evaluating images generated by this version, some of which you can see below. Just keep in mind that since the default resolution is 1024×1024, the Midjourney V5 generator may take a bit longer to generate creations than its predecessor, but you’ll immediately be able to notice the difference.
- Striking 10.9-inch Liquid Retina display with True Tone
- A14 Bionic chip with 6-core CPU and 4-core GPU
- 12MP Wide back camera
This remarkable feat of technological advancement is not to be taken lightly, my friends. The V5 release boasts significant improvements in language understanding, accuracy, and aesthetic flexibility, all of which contribute to the generation of images that are nothing short of breathtaking,” said Eva Rtology, Midjourney Artist.