Hehehe, don't worry, this time it's not text-to-image generation. I know many prompts are long and difficult to write, and writing a lot of them can be exhausting.

Therefore, this flow is specifically for image-to-image scenarios.

However, it’s not purely image-to-image processing. Strictly speaking, it’s a combination of redrawing and transformation. If you upload a 2D illustration, it will transform it into a realistic image.

If you upload a realistic image, it will transform it into a cartoon-style image.

What does this remind you of? Bingo! Style transfer! But this flow is a bit different. Style transfer aims to preserve the details of the original image, which this does not. The resulting image will still look like it’s been reprocessed.

Rather than calling it style transfer or style migration, it’s more accurate to call it a combination of style transformation and image processing.

If you want to interfere with the output process, such as specifying that the resulting image should have an ink painting style or a Studio Ghibli style, you can write the instruction in the additional input box. Note that using English is the best choice for the instructions.

As for specific use cases, that’s up to the users to explore on their own~



Recommended image dimensions:
Best for portraits (vertical orientation):
1024 x 1280, 896 x 1344

General purpose:
1024 x 1024, 1344 x 896, 1344 x 768

Avoid:
1080 x 1350, 1536 x 2048



Additionally, there’s a text-to-image version of the application next door. You can input creative ideas and directly generate images. For details, please see: https://www.runninghub.cn/ai detail/2070211417835397122

————————Finally, the usual advertisement:
New users who don’t have enough points can register using my link to get 1000 RH coins: https://www.runninghub.cn/?inviteCode=1a9e6e17
Users who are already registered but haven’t entered an activation code can also use my activation code to receive 1000 points: 1a9e6e17