SenseMirage by SenseTime is an AI content creation community platform launched by SenseTime Technology in China. Based on SenseTime's self-developed AIGC model, SenseMirage has over 1 billion parameters and supports various generation styles such as two-dimensional and three-dimensional.
Recently, SenseMirage has upgraded its features, especially in terms of being free and efficient, showing impressive performance. This article will provide a detailed introduction to the main features and user experience of SenseMirage.
Introduction to SenseMirage#
SenseMirage is an AI content creation platform launched by SenseTime Technology. Users can generate high-quality images in a very short time by entering prompts. The maximum resolution can reach 6K. SenseMirage supports various generation styles and provides rich AIGC model resources.
- Website: https://miaohua.sensetime.com/
Features of SenseMirage#
1. Super-fast Image Generation#
The strength of SenseMirage lies in its image generation speed. SenseTime's self-developed painting model can generate four images with a resolution of 1536*2048 in 20 seconds, greatly improving the efficiency of image generation.
2. Technological Advancement#
SenseMirage has advantages in text understanding, diverse image styles, and high-quality detail generation due to SenseTime's core technology algorithms. Users can use official models or choose community open-source models to complete their artwork.
3. Model Marketplace#
SenseMirage is not only a creative tool but also provides rich model resources, including self-developed models by SenseTime and third-party community open-source models. Users can choose or train customized LoRA models according to their needs to generate personalized content.
4. LoRA Model Training#
Users can upload local images and train their own LoRA models using SenseTime's self-developed or open-source models to generate personalized content. Reference
Instructions for Using SenseMirage#
After registration, users can enter the AI drawing interface. The interface and operation of SenseMirage are similar to SD (Stable Diffusion), but they also incorporate the advantages of Midjourney. Here are the steps to use it:
Selection of Large Models#
SenseMirage provides five large models for users to choose from. Clicking on the model details will display the introduction, but the first three models do not support Lora and ControlNet models.
Lora and ControlNet#
On the left side, users can choose models, Lora and ControlNet control models, and perform operations such as image generation and local redraw. Currently, there are fewer Lora models, but custom training is possible.
Positive and Negative Prompts#
On the right side, there is a prompt setting area, including positive prompts and negative prompts. Positive prompts are used to describe image styles and details, while negative prompts are used to avoid generating low-quality or unnecessary elements.
Resolution and Number of Generated Images#
Users can choose the aspect ratio and resolution of the images, and up to 8 images can be generated simultaneously. The generation of 2K resolution images can be completed within 10 seconds.
Inspiration Plaza#
In the Inspiration Plaza, users can view the works of other players and generate similar style images through the one-click generation function. This is a good way to get inspiration and quickly generate images.
Model Plaza#
The Model Plaza brings together official and community-trained models. Users can choose different models for creation, and the number of models is constantly increasing.
Generated Image Showcase (AI-generated prompts)#
On the SenseMirage platform, you can try image generation in various styles and themes. Here are some innovative examples of image creativity, showcasing the unique expression of SenseMirage in different styles.
Fantasy Flowers#
Prompt: A mysterious night-blooming flower with petals that shine like colorful silk, surrounded by faint lights and twinkling stars. The background is a dark night, and the flower's light penetrates the darkness, illuminating the surrounding grass, like a dreamy garden feast.
Futuristic Portrait#
Prompt: A warrior wearing futuristic armor, with half of the face hidden behind a transparent mask. The eyes emit sharp light. The background is a minimalist futuristic city skyline. The whole picture is filled with high-contrast light and shadow effects and a cold blue tone, showing the coolness and mystery of the future world.
Poetic Landscape#
Prompt: A landscape painting that combines traditional Chinese ink painting and modern minimalist style. Distant mountains are layered, and the lake is like a mirror with gentle ripples. The foreground consists of lush bamboo forests and floating white mist, creating a surreal sense of tranquility, as if entering an ancient poetic painting.
Fairy Tale Animal#
Prompt: An adorable rabbit wearing a gorgeous ball gown, standing on a decorated stage. The surroundings are filled with dazzling lights and colorful fireworks. The rabbit's eyes sparkle with magical radiance. The whole scene is full of fantastic fairy tale colors and a festive atmosphere.
Sci-Fi Future Scene#
Prompt: In an alien city far away from Earth, under a huge transparent dome, glowing flying vehicles float in the air. The streets are filled with strange plants and futuristic-style buildings. The distant starry sky is magnificent, presenting a grand and peculiar alien landscape.
Magical World#
Prompt: A witch wearing a gorgeous magic robe casting a spell. The wand in her hand emits dazzling light. Rotating magic runes and floating stardust surround her. The background is an ancient magic tower. The whole scene is full of mysterious magic and fantastic light effects.
Portrait Photography#
Prompt: A young Chinese woman in her twenties with a round face, beautiful and clear facial features, a lovely appearance, a well-balanced and proportionate face shape, pure and natural bare face, transparent and fair skin, three-dimensional facial features, a star-like face, upper body photo, passport photo, front view, long hair.
Anime Drawing#
Prompt: A girl wearing a gorgeous battle suit, standing in a futuristic city background. Her eyes are full of determination and courage. The background has dazzling light effects and detailed architectural details, presenting a vibrant and colorful anime-style scene.
Summary of Use#
Overall, the image quality generated by SenseMirage is relatively high and can be compared to Midjourney in some usage scenarios. As the name "SenseMirage" suggests, its greatest advantage is its fast generation speed. However, some portraits drawn with SenseMirage still lack realism, have deviations in composition, and have low clarity. It is expected that SenseMirage will make greater progress in model iteration and improvement in the future. Currently, after registration, users can generate 10 images per day, but by inviting friends to register, both parties can enjoy three days of unlimited generation. Interested friends can register and experience it through the following link: