Your Face In Midjourney! It's Here, Easy, and It Works! (Amazing Faceswap Plugin!)

Theoretically Media
18 May 202309:48

TLDRIn this video, Tim introduces an innovative plugin that allows users to seamlessly swap their faces into mid-journey images without the need for complex editing software like Photoshop. He demonstrates the process by creating a server in Discord, adding the Insight face and Mid Journey bots, and then using his own face to generate images. Tim showcases the plugin's capabilities by transforming his face into various characters and styles, including a cinematic still of Indiana Jones, a painting in the style of Andrew Wyeth, and a comic book character. While noting some limitations, such as issues with beards and extreme expressions, he emphasizes the plugin's potential and the ongoing development by the Insight team to improve its functionality. Tim concludes by encouraging responsible use of the face swap feature.

Takeaways

  • 😃 With the new InsightFace plugin, users can now swap their face into Midjourney images without relying on Photoshop or uploading selfies.
  • 📸 The plugin works directly in Discord, and you need to set up a server to get started. The Midjourney bot is also necessary for the image generation.
  • 📋 You can save your face ID by running a command and providing a forward-facing, high-quality photo without glasses or bangs.
  • 🎥 The plugin can produce cinematic and artistic results, turning the user's face into iconic characters like Indiana Jones or Superman.
  • 🎨 Different artistic styles, such as those of Andrew Wyeth, Wes Anderson, or Darwin Cook, can be utilized for unique image generation.
  • 🤖 Some stylized images may not fully replicate the intended style but can still offer a remarkable resemblance.
  • 👨‍💼 Users can create professional photos, such as LinkedIn profile pictures, or avatars for business development.
  • 💪 The plugin struggles with specific challenges like extreme facial expressions, complex lighting, and facial hair.
  • 📈 As an evolving tool, future updates and features are expected to further enhance the plugin's performance.
  • 🚫 Users are reminded to swap faces responsibly and consider ethical implications when utilizing this technology.

Q & A

  • What is the main purpose of the 'Inside Face' plugin discussed in the script?

    -The 'Inside Face' plugin allows users to swap their faces into mid-journey images without the need for advanced photo editing software like Photoshop.

  • How does the user create a new server in Discord for the plugin?

    -The user creates a new server in Discord by clicking the 'add a server' button, creating their own server, naming it, and then hitting 'create'.

  • What is the first step to use the 'Inside Face' plugin after adding it to the server?

    -The first step is to create an ID for the face to be swapped by using the 'forward slash save ID' command and uploading an image of the face.

  • What are some limitations of the 'Inside Face' plugin when handling certain styles or expressions?

    -The plugin struggles with extreme expressions, stylized or abstract images, and may have issues with beards or glasses. It also doesn't re-roll images, so repeated attempts will yield the same result.

  • What kind of image should be used for the 'Inside Face' plugin to work effectively?

    -The image should be a high-quality, forward-facing profile picture without glasses and without bangs, as these can interfere with the face-swapping process.

  • How long does it typically take for the 'Inside Face' plugin to generate an image?

    -The plugin can generate an image in about 40 seconds or less, depending on the complexity of the request.

  • What happens when you try to swap a face with a beard using the 'Inside Face' plugin?

    -The plugin may struggle with beards, sometimes resulting in a distorted or incorrect representation of facial hair.

  • What does the user suggest doing if the initial image from the 'Inside Face' plugin isn't satisfactory?

    -The user suggests doing re-rolls to get a better base image and possibly merging images in Photoshop for a more seamless result.

  • How does the 'Inside Face' plugin handle more artistic styles, such as an Andrew Wyeth painting?

    -The plugin can generate images that are in line with the requested artistic style, and when a face swap is performed, it maintains the style consistency.

  • What is the user's opinion on the final result of the face-swapped images?

    -The user finds the results to be of good quality from a technical standpoint, with consistent skin tones and lighting, although some results are slightly disturbing to him personally.

  • What does the user recommend for responsible use of the 'Inside Face' plugin?

    -The user recommends using the plugin responsibly and being aware of its limitations, as well as expecting improvements as the developers continue to work on it.

Outlines

00:00

🌟 Introduction to Face Swapping with Mid-Journey

This segment introduces an easy method to integrate your face into Mid-Journey images, eliminating the need for Photoshop or repeated attempts. The speaker details the process starting from creating a new server on Discord, adding the necessary bots for Mid-Journey and Insight Face, and setting up an ID for face swapping. Examples are given using the speaker’s face superimposed onto characters like Indiana Jones and Superman, demonstrating the integration's simplicity and effectiveness. The segment also highlights requirements and tips for selecting the right photos to ensure quality results in the face swapping process.

05:00

🎨 Exploring Advanced Face Swapping Techniques and Challenges

This section delves into more complex scenarios using the face swapping feature, testing its limits with different styles and settings. The speaker describes successfully merging their face with various artistic styles, including a painting mimicking Andrew Wyeth and an 80s cyberpunk hero illustration. However, challenges are noted, such as color tone inconsistencies and issues with facial hair integration in more detailed or stylized images. The segment further discusses the limitations of face swapping in extremely stylized or abstract visuals like comic book styles, and ends with practical applications, showcasing how the technology can be used to create professional-looking images for platforms like LinkedIn.

Mindmap

Keywords

💡Faceswap

Faceswap is a technology that involves replacing the face of a person in a photograph or video with another person's face. In the context of the video, the term is used to describe the process of superimposing the user's face onto various characters or images using the Insight Face plugin, which is a significant part of the video's theme.

💡Mid-journey

Mid-journey refers to a specific point in a process or journey. In the video, it is used as the name of a bot that is capable of generating images based on textual prompts. It plays a central role in the video as the tool that creates the base images into which the user's face is swapped.

💡Insight Face

Insight Face is an open-source library for 2D and 3D face analysis. It is used in the video to facilitate the face-swapping process by allowing users to upload an image of their face and then apply it to different images generated by the Mid-journey bot. It is a key technology demonstrated in the video for its ability to transform user-supplied images.

💡Discord

Discord is a popular communication platform that supports text, video, and voice conversations. In the video, it is used as the platform where the user interacts with the Mid-journey and Insight Face bots to create and manipulate images. It serves as the environment in which the face-swapping process takes place.

💡Aspect Ratio

Aspect ratio is the proportional relationship between the width and the height of an image or screen. In the video, the term is used to describe the dimensions of the generated images, such as 16:9, which is a common aspect ratio for widescreen displays. It is an important parameter when creating images that are intended for specific display formats.

💡Cinematic Still

A cinematic still refers to a single, static frame extracted from a movie or video that has the qualities of a still photograph. In the video, the user requests 'cinematic still' images of various characters and scenes, which are then used as the basis for the face-swapping process, emphasizing the video's focus on creating visually appealing and realistic image transformations.

💡Superman

Superman is a well-known fictional superhero character. In the video, the user demonstrates the face-swapping technology by placing their face onto an image of Superman, which serves as an example of how the plugin can be used to fulfill personal fantasies or create novel images.

💡Wes Anderson

Wes Anderson is a renowned film director known for his distinctive visual and narrative style. The video references 'Wes Anderson film' to describe a prompt given to the Mid-journey bot, which generates an image in the style of Anderson's movies. This keyword highlights the video's exploration of the plugin's ability to mimic specific artistic styles.

💡Upscale

To upscale an image means to increase its resolution or size without losing quality. In the video, the user upscales the generated images to improve their detail and clarity before applying the face-swapping process. It is a step that enhances the final output of the face-swapping plugin.

💡Andrew Wyeth

Andrew Wyeth was a famous American realist painter known for his detailed and emotive rural landscapes. The video mentions 'painting style by Andrew Wyeth' when the user prompts the Mid-journey bot to create an image in the style of Wyeth's painting 'Christina's World'. This keyword showcases the versatility of the plugin in capturing different artistic styles.

💡Cyberpunk

Cyberpunk is a genre of science fiction that features advanced technological and scientific achievements, juxtaposed with a degree of breakdown or radical change in the social order. In the video, the user experiments with a 'cyberpunk hero' prompt to test the face-swapping plugin's capabilities with more stylized and abstract imagery, demonstrating the plugin's potential for creating images in various fictional genres.

Highlights

A new plugin allows users to swap their faces into mid-journey images without the need for Photoshop or multiple uploads.

The process is done entirely within the mid-journey platform, making it user-friendly and efficient.

The plugin uses the Insight Face open-source library for 2D and 3D face analysis.

Creating a new server on Discord is the first step in using the plugin.

The user is guided through adding the plugin to their server with a simple process.

The Mid Journey bot must be invited to the server to enable the use of slash commands.

Users can create an ID for the Insight bot by uploading an image and naming their character.

The image supplied to Insight Face should be high-quality, forward-facing, and without glasses or bangs.

The plugin can generate images in various styles, including cinematic stills and detailed paintings.

The face-swapping feature works well for realistic and photographic styles, maintaining skin tones and lighting consistency.

The plugin has limitations with highly stylized or abstract images, and may struggle with extreme expressions.

The Insight developers are actively working on improvements, suggesting future updates may enhance functionality.

The plugin can be used to create convincing LinkedIn profile pictures with some re-rolls and base image adjustments.

Beards and glasses can sometimes cause issues with the face-swapping algorithm.

The plugin does not re-roll images, so repeated attempts will result in the same output.

The user, Tim, demonstrates the plugin's capabilities by transforming into various characters and settings.

The plugin offers a unique and fun way to fulfill childhood fantasies and explore different personas.

Users are encouraged to use the face-swapping plugin responsibly.