Hello everyone, I am excited to introduce the launch of NFTs by A2SET.
You can find the NFT works of A2SET VISION on Opensea. A2SET-VISION is a collection of works created through a complex combination of various technologies.

A2SET VISION
The ‘A2SET_VISION’ collection represents a fusion of AI technology and artistic flair, manifesting as a digital art piece centered around a woman.
It is a creation born from an extensive database, learning various styles and featuring customized skin, facial features, and hair.
Beyond the surface-level visual appeal, this artwork encapsulates a deeper significance β every element is composed of data, combining creativity with AI to create a unique overall mood. Now, let’s embark on the journey to explore this series.

Description
Essential for introducing any artwork is its accompanying description! As a platform dealing with AI, we have given A2SET VISION a simple yet powerful form of explanation. We have unraveled the Description in the form of Positive Prompts, which are all too familiar to those who have experienced the AI Art Generator.
We have explained the manual using prompts that are not too long and only cover the most symbolic elements, further emphasizing the unique characteristics of A2SET VISION.
Traits
The Traits of A2SET VISION can range from as few as 2 to as many as 4 in creation.
- HAIR COLOR
- JACKET COLOR
- MOOD
- BACKGROUND
Please note that all future A2SET VISION artworks will be created based on these Traits.
How is A2SET VISION created?
A2SET VISION is created through a three-stage process with additional steps
1. AI Prompt Process
Initially, a general framework is produced using Girl Face Data created by A2SET, through Stable Diffusion, Midjourney and Leonardo AI.
As those who have used AI Art Generator know, we used prompts and various settings to create the female figure for the A2SET VISION collection.

As you can see above, through the combination of A2SET’s flagship product models and training with 3D Stylized Mesh, we have created a casual look that mixes Western and Eastern styles.
- Some parts were created through base model training, while others were made using LoRA training.
- –> Go to Post: Learn about LoRA Training
2. AI Joint & Line Process
For refining the processed images, we utilize ControlNet’s specific functions with customized images. We create concise yet beautiful images using source materials specific to Openpose for reflecting joint values in the face, body, and fingers, and A2SET’s own drawing images for Soft Edge and Line Art.

For instance, while maintaining a single Openpose image that represents the same pose, we create specific drawing and alpha values for harmonizing Soft Edge, Line Art, and prompts. The images produced through the merging and computation of these various data sets can be considered the final step before our editing process.
The most crucial aspect here is the drawing part. We determined that there are limitations to expressing the desired clothing styles using prompts alone. Therefore, by undergoing a more intuitive and separate drawing process to obtain a variety of Partial data, we were able to produce more precise results.
3. Editing Process
The A2SET VISION series is created by conducting additional editing work on the pattern data, which includes our brand logo, and reflecting a simple background to present a clear output.

This means that A2SET VISION is created through a multitude of databases and processes. Of course, it’s not possible to detail every single process in this article, but it would be appreciated if you understand that each piece is carefully crafted based on the key processes mentioned above.
* As a note, a Christmas edition commemorating Christmas 2023 has also been minted, so those interested, please enjoy π

We plan to continue presenting NFT artwork through the A2SET VISION collection, so your interest and support would be greatly appreciated.