Consistent Characters Soon | Midjourney Office Hours Recap Feb 28th 2024 | Midjourney News

Future Tech Pilot
29 Feb 202403:25

TLDRThe Mid Journey office hours recap highlights the team's progress on three main features: a new character reference, style reference system, and describe tool. They are currently testing the character reference and plan to launch all features soon. The updates aim for better accuracy and style transfer, with potential moderation for realistic images. The team is also exploring a random style generator and improving the mobile experience, with possible API openings for chatbot companies. Server upgrades are expected to reduce image generation time.

Takeaways

  • 🎉 The Mid Journey team is actively working on three main features: a new character reference feature, a new style reference system, and a new describe feature.
  • 🚀 These features are expected to launch soon, with testing currently underway for two versions of the character reference.
  • 🌟 Character reference might allow for anime characters to appear in real-life photography, showing excitement for the potential of blending styles.
  • 📈 The process of adding these features is seen as a significant step forward, with continuous improvement expected in version 7.
  • 🔍 Efforts are being made to refine the ability to capture smaller details, though some challenges remain in pulling out these elements.
  • 🪂 New moderation measures may be introduced, potentially starting as strict and then adjusted based on further understanding and feedback.
  • 🎨 The style reference update will replace the old system, aiming for more accurate style transfer, despite breaking the previous one.
  • 🤖 Discussions about a random style generator and a style map have been mentioned, suggesting future exploration of style space and discovery.
  • 🚀 Server upgrades are in progress to enhance speed, aiming to reduce image generation time from 35 seconds to 15-20 seconds.
  • 📱 The social aspect of the Mid Journey website is being worked on, with a focus on bug fixes and improving the mobile user experience.
  • 🔌 The team is considering opening up an API for certain chatbot companies, hinting at potential collaborations in the future.

Q & A

  • What is the main focus of the MidJourney team at the moment?

    -The MidJourney team is currently working on three main features: a new character reference feature, a new style reference system, and a new describe feature.

  • What are the two versions of character reference being tested?

    -The exact details of the two versions are not specified in the script, but it is mentioned that they are testing different aspects of the character reference feature.

  • What is the potential limitation of the character reference feature?

    -The character reference feature might not capture small details like the outfit a character is wearing or minor facial features. However, it could allow anime characters to appear in real-life photography.

  • How does David view the development process for these new features?

    -David sees the development process as a big step forward and believes that the features will improve in future versions, specifically mentioning that version 7 is expected to be even better.

  • What is the issue with the style reference update?

    -The style reference update will break the old system, but it will be more accurate at transferring the style. There have been some issues with links on Discord related to image prompting problems.

  • What challenges are there in achieving better body coherence?

    -Achieving better body coherence has been challenging for the team, as they are trying to push the limits of version 6 without hitting diminishing returns.

  • What is the potential new feature after the current three are launched?

    -After the current three features are launched, the team is considering working on a random style generator, similar to SRF random, which would allow users to explore the style space better.

  • How will version 6 improve in terms of speed?

    -Version 6 is expected to get faster as the team is upgrading their servers. They hope to reduce the time to generate an image from 35 seconds to 15 or 20 seconds, in addition to introducing a turbo mode in the future.

  • What is being done to improve the social aspect of the MidJourney website?

    -The team is currently working on fixing bugs and improving the mobile experience for the website's social features.

  • Is there a plan to open up an API for MidJourney?

    -The team is exploring the possibility of opening up an API for some chatbot companies, although they did not specify which companies.

  • How can viewers see more of the art featured in the video?

    -Viewers can follow the creator on Instagram @futuretechpilot to see more of their artwork and illustrations.

Outlines

00:00

🎥 Mid-Journey Office Hours Recap

This paragraph recaps the Mid-Journey office hours held on February 28th. The video is sponsored by patrons who support the creator on Patreon for $5 a month, gaining access to multiple prompts weekly. The Mid-Journey team is focused on three main developments: a new character reference feature, a style reference system, and a new describe feature. They are aiming to launch all three soon, with current tests on two versions of the character reference and plans to test the describe feature soon. The character reference may include anime characters in real-life photography, with potential moderation for positive imagery. The new describe feature is expected to be more accurate, and the style reference update may break the old one but will offer better style transfer. Discord link issues and body coherence challenges are being addressed. The team is excited about version 7, and server upgrades may speed up version 6. The social aspect of the website is being improved, and there are plans to open an API for certain bot companies.

Mindmap

Keywords

💡Mid Journey

Mid Journey refers to a specific phase or version in the development process of a project or software. In the context of the video, it relates to the ongoing work and updates being made to an AI-based art generation platform. The term is central to the video's theme as it signifies progress and the continuous improvement of the platform's features.

💡Office Hours

Office hours typically refer to the scheduled time when a professional, such as a developer or a company representative, is available to interact with the public or users. In this video, 'Office Hours' is used to describe a recap of a Q&A or update session where the team behind the AI art platform shares their progress and plans with the community.

💡Character Reference

Character Reference in the context of this video pertains to a new feature being developed for the AI art platform that will allow users to input references for creating characters. It is an essential aspect of the platform's functionality, as it helps in generating more accurate and user-desired character representations in the artwork.

💡Style Reference System

Style Reference System refers to an upcoming feature that aims to improve the AI's ability to transfer a specific artistic style onto generated images. This concept is crucial for users seeking to have more control over the aesthetic outcome of their AI-generated art, allowing them to achieve a desired style more precisely.

💡Describe

In the video, 'Describe' is mentioned as a feature that the team is working on to enhance the accuracy of the AI in generating images based on textual descriptions. It is a key component in the user experience, as it directly affects the relevance and quality of the AI-generated content.

💡Anime Character

Anime Character is used in the script to illustrate a potential application of the character reference feature, where an anime-style character could be integrated into real-life photography. This example showcases the versatility and creative possibilities that the AI art platform is aiming to offer its users.

💡Moderation

Moderation in this context refers to the oversight and management of content to ensure it adheres to certain guidelines or standards. The video discusses the potential need for moderation in the form of positive imagery requirements for realistic depictions of people, highlighting the platform's commitment to maintaining a safe and respectful environment for its users.

💡Body Coherence

Body Coherence is a term used in the video to describe the challenge of generating AI art that accurately represents the human form in a cohesive and anatomically correct manner. The platform is working on improving this aspect to enhance the quality and realism of the generated images.

💡Style Map

Style Map is mentioned as a concept for a tool that would allow users to explore the 'style space' more effectively, potentially finding and selecting references that they like for their AI-generated art. This concept relates to the platform's goal of providing users with more control and precision in styling their creations.

💡API

API, or Application Programming Interface, is a set of protocols and tools that allows different software applications to communicate with each other. In the video, the mention of exploring an API opening suggests the platform's intention to enable third-party developers to create applications or bots that can interact with the AI art platform, expanding its capabilities and user base.

💡Server Upgrade

Server Upgrade refers to the process of improving the hardware or software of the servers that host a platform or service. In this video, the server upgrade is aimed at reducing the time it takes to generate images, from 35 seconds to 15 or 20 seconds, enhancing the user experience by making the platform faster and more efficient.

Highlights

The Mid Journey team is working on three main features: a new character reference feature, a new style reference system, and a new describe feature.

The team aims to launch all three features next week.

They are currently testing two versions of the character reference feature.

Small details like outfits or facial features might not be captured in character references, but larger, more general themes like an anime character in real-life photography could be.

The process is seen as a big step forward, with potential for further improvements in version 7.

Moderation updates may be introduced, requiring positive imagery, particularly for realistic depictions of people.

The new describe feature is expected to be more accurate and get users closer to the source image.

The style reference update will break the old system but promises a more accurate style transfer.

There have been issues with Discord links and image prompting problems, but they are actively being addressed.

Efforts are being made to improve body coherence, although it has been a challenging task.

Version 7 is being eagerly anticipated, with no signs of diminishing returns in sight.

A random style generator is being considered as the next feature after the current three.

The idea of a style map for better exploration of style space is being discussed.

Version 6 is expected to become faster with server upgrades, potentially reducing generation time from 35 seconds to 15-20 seconds.

Plans for a turbo mode in the future are also being considered.

The social aspect of the website is being worked on, with a focus on fixing bugs and improving the mobile experience.

The team is exploring the possibility of opening up an API for certain chatbot companies.

The art in the video is created using a prompt from the Mid Journey website, showcasing a variety of styles and techniques.