Image Source: Generated by Wujie AI
It has been half a year since the release of Midjourney v5. When will the next major version v6 be released? The answer is: later this year.
David Holz, CEO of Midjourney, stated at an event yesterday that the leap from v5 to v6 will be greater than from v4 to v5. However, Holz did not provide an exact release date, but indicated that it will be soon, definitely within this year.
Regarding what new features Midjourney v6 will bring, Holz promised to bring better text understanding, meaning that image generation will be closer to the described prompts and will better capture the details in the prompt wording. It may also bring text rendering functionality.
Holz said these are "not difficult." However, he also mentioned that he has not seen very good text rendering and remains open to whether this feature will be included and to what extent it will be effective.
OpenAI Challenges Midjourney with V6
Last week, OpenAI released a new version of its image generation model, DALL-E 3. DALL-E was one of the earliest generative AI models, predating Midjourney. The DALL-E 3 version has seen significant improvements in various aspects. Many even believe that DALL-E 3 will pose a direct threat to Midjourney and has already surpassed Midjourney in some aspects, especially in its understanding of prompts and details after being combined with ChatGPT.
When asked about the release of OpenAI's DALL-E 3, Holz "optimistically" believes that Midjourney will continue to provide the highest image quality. The initial comparison between DALL-E 3 and Midjourney v5 shows that DALL-E 3 is not significantly ahead in terms of image quality, but it does better follow prompts and can render text.
What's New in Midjourney v6?
3D and Video
Midjourney's roadmap also includes the functionality to create 3D images and videos. Especially for 3D images, Holz is "optimistic" and believes that there will soon be a turning point.
When he looks at current video games, he expresses surprise at the poor quality of image and the significant contribution of generative AI to quality. Holz has previously stated that he expects future video games to be generated, not rendered. The 3D image functionality will be a boon for VR and AR game creators.
Midjourney does not intend to release any specific information or demos about 3D generation this year. The same goes for the video generation being developed by the Midjourney team, but Holz mentioned that video generation may be later and is not yet ready.
Higher Resolution
One of the current major drawbacks of Midjourney is the relatively low resolution of the generated images. The default image size is currently 1,024 x 1,024 pixels. Other aspect ratios are available, but they do not increase the resolution of 72ppi. For example, the resolution of a 16:9 image can only reach 1,456 x 816 pixels.
This resolution is generally suitable for web publishing but is not sufficient for any large-scale printing. Therefore, support for higher resolutions is an urgent user demand that needs to be addressed.
Other Updates
New Niji Style
The Midjourney team is also developing a new Niji style. The previous two-dimensional model received unanimous praise from anime enthusiasts upon its launch, and the new Niji model is expected to bring a lot of upgrades and improvements.
New Website
The web version of Midjourney is still under development. The new web version will be rolled out in two stages:
First stage: Ability to view, search for one's own images, etc.
Second stage: Will have image creation and social features
However, Holz did not provide a specific timeline for the launch.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
