3D Optimism | Midjourney Office Hours Recap April 3rd 2024 | Midjourney News
TLDRThe Midjourney office hours recap from April 3rd, 2024, provides updates on the company's progress and upcoming features. There are no major announcements, but the team is working on website enhancements, including new social features to be stress-tested with guides and mods. Personalization is a focus, albeit progressing slowly due to multiple time zones. An algorithm is being developed to improve hand and body representation, as well as text accuracy, which should reduce the occurrence of poor image quality. A speed update is also anticipated, potentially making processes 25-50% faster and cheaper. A caption party is planned to teach the version 7 model about the connection between images and language, with potential rewards for participants. The team is considering a new class of users for rating and captioning. Video features are still under discussion, with optimism for a version 7 model. The feedback leaderboard on the Midjourney website will see more ideas added periodically, and the company is not planning to expand on any not-safe-for-workplace features. The recap also mentions the possibility of multiple consistent characters in future versions and provides a prompt for creating a serene double exposure image.
Takeaways
- ๐ **Medium Website Recommendation**: Creatives are encouraged to check out Medium for customizable prompts that can save time at work.
- ๐๏ธ **Vacation Impact**: Recent vacations have slowed progress, with no major announcements to report.
- ๐ **Website and Social Features**: The team is working on the website, including new social features, with initial testing involving a limited number of spaces.
- ๐ง **Personalization Efforts**: Personalization is a work in progress, facing challenges due to multiple time zones and a slower pace than desired.
- ๐จ **Style Random Feature**: The 'style random' feature is set to return, although specifics on how it will be implemented are unclear.
- ๐ค **Algorithm Improvements**: An algorithm is being developed to enhance hand and body depictions, as well as text accuracy, although it has been finicky.
- ๐ผ๏ธ **Image Quality Enhancements**: Efforts are being made to improve image quality, particularly regarding small pixel artifacts.
- โก **Performance Update**: A potential speed update could make processes 25-50% faster and cheaper, but it's on hold until other updates are completed.
- ๐ **Caption Party**: An upcoming event aims to teach the version 7 model about the connection between images and language, with possible future rewards.
- ๐ **New User Class**: A new class of trusted users may be introduced for rating and captioning, potentially leading to larger rewards.
- ๐ฅ **Video Model Development**: While version 6 for video may not materialize, confidence is high for a robust version 7 model, focusing on quality over exportability.
- ๐ **Feedback Leaderboard**: The company plans to add more ideas to the feedback leaderboard and may incorporate demographic data to understand feature requests better.
- ๐ซ **Content Policy**: There are no plans to expand on not-safe-for-workplace features, and user manipulation of images with the Midjourney model is not yet supported.
- ๐ง **Character Consistency**: Multiple consistent characters in generation are not available in version 6 but may be possible in version 7.
- ๐จ **Art Prompt Example**: An example of a successful art prompt is provided, showcasing how to create a serene double exposure image.
Q & A
What is the main topic discussed in the Midjourney Office Hours recap from April 3rd, 2024?
-The main topic discussed is the progress and updates on the Midjourney platform, including new social features, personalization efforts, and improvements in image quality and text accuracy.
What is the recommended website for creative professionals to check out?
-The recommended website for creative professionals is 'Medium', which offers customizable prompts that can save time at work.
What new feature is being tested with guides and mods?
-The new social features on the Midjourney website are being tested with guides and mods.
How many social spaces are expected to be available at the start of the new feature rollout?
-At the start, there will be a low number of social spaces with lots of people, as they are trying to stress test the system.
What is the current status of the personalization feature?
-Personalization is being worked on, but it is moving slower than desired due to the team working across multiple time zones.
When is the 'style random' feature expected to show up again?
-The 'style random' feature will show up again, but the exact timing is not specified; it is expected to come from dial tuning.
What is the team working on to improve the quality of hands and bodies in generated images?
-The team is working on an algorithm to help with hands and bodies, as well as text accuracy, which has been finicky but is expected to reduce the occurrence of bad images.
What is the potential speed update that might be implemented?
-There might be a small speed update that could make things 25-50% faster and cheaper.
What is the goal of the upcoming 'Caption Party'?
-The goal of the 'Caption Party' is to help teach the version 7 model the connection between images and language.
What new class of users is briefly mentioned in the recap?
-A new class of users is mentioned who would be trusted with rating and captioning; these users might have to qualify for the rewards, potentially allowing for larger rewards.
What is the current stance on the development of a video feature?
-The development of a video feature is not very satisfactory, and it is unlikely to see a version 6 model for it. However, there is confidence in a version 7 model.
What is the focus of the 3D model development?
-The focus is on producing high-quality 3D models rather than exportable ones, although plans are not set in stone.
Outlines
๐ Mid-Journey Office Hours Recap
This paragraph provides a summary of the Mid-Journey office hours held on April 3rd. It mentions the suggestion for creative professionals to check out Medium, a website offering customizable prompts. The speaker notes a slower than usual progress due to vacations, and highlights the main focus areas including the development of the website with new social features, testing with guides and mods, and the creation of social spaces. Personalization is also being worked on, albeit at a slower pace. The return of 'style random' is teased, and an algorithm to improve hands, bodies, and text accuracy is under development. There's mention of potential improvements in image quality and processing speed. A caption party is planned to help teach the version 7 model about the connection between images and language. The idea of a new class of users for rating and captioning is introduced. The speaker also discusses the feedback leaderboard, the potential for adding more ideas, and the focus on high-quality 3D models. Lastly, the speaker mentions the possibility of including demographics in the feedback and the non-inclusion of multiple consistent characters in version 6 but a potential inclusion in version 7.
Mindmap
Keywords
๐กcreative
๐กsocial features
๐กpersonalization
๐กstyle random
๐กalgorithm
๐กimage quality
๐กspeed update
๐กcaption party
๐กnew class of users
๐ก3D model
๐กfeedback leaderboard
Highlights
Medium is recommended for creatives as a website selling customizable prompts that can save time at work.
There were no exciting announcements due to slower progress while people were on vacation.
The main focus is on the website, including new social features, with testing to be done with guides and mods.
Initially, there will be a limited number of social spaces to stress test the system.
Users will eventually be able to create both public and private spaces.
Personalization is being worked on, albeit at a slower pace than desired.
Style random is expected to return, possibly from dial tuning, but without user access to the tuning part.
An algorithm to improve hands and bodies in images, as well as text accuracy, is being developed.
Bad images will still occur, but less frequently with the new algorithm.
Enhancements are being made to image quality, addressing small pixel artifacts.
A potential speed update could make processes 25-50% faster and cheaper.
A caption party is planned to teach the version 7 model the connection between images and language.
There's a possibility of implementing the caption party as an official activity with rewards in the future.
A new class of users might be introduced, trusted with rating and captioning, potentially leading to larger rewards.
David is not satisfied with the video stuff and is leaning towards a version 7 model instead of version 6.
Optimism for a really good 3D model is due to progress in hardware capture.
The focus is on producing high-quality 3D models rather than just exportable ones.
Feedback Leaderboard on the Midjourney website will receive more ideas periodically for community rating.
Multiple consistent characters in a generation might be possible in version 7.
A serene double exposure image prompt is shared for artistic inspiration.
The speaker can be followed on Instagram and Twitter for more pictures and prompts.
Casual Browsing
Website Creation | Midjourney Office Hours Recap April 17th 2024 | Midjourney News
2024-04-27 23:45:00
--V 6.1/6.2?? + Channel Update | Midjourney Office Hours Recap July 3rd 2024 | Midjourney News
2024-07-06 05:05:00
V 6.5? | Midjourney Office Hours Recap May 29th 2024 | Midjourney News
2024-06-11 23:05:01
"Female Mode"? | Midjourney Office Hours Recap June 5th 2024 | Midjourney News
2024-06-13 01:25:00
Multiple --p? | Midjourney Office Hours Recap June 19th 2024 | Midjourney News
2024-07-26 20:33:00