The AI startup Higgsfield has recently captured significant attention with its cutting-edge Soul AI image generation software, enabling users to produce breathtaking images using a single text input. While generative AI tools are not a novelty, with OpenAI’s ChatGPT and Google’s Gemini providing comparable functionalities, Soul sets itself apart with a distinctive capability that makes its images look like they were taken with an iPhone. The company has crafted more than 50 visual styles for users to select from prior to entering their text prompt, facilitating the generation of images that seem as if they were taken with a camera.
A crucial aspect of any AI image generator is the functionality to modify the images it produces, and Higgsfield has launched an Inpaint feature to serve this need. Currently in beta, this feature allows Soul users to easily edit particular elements of an image. Publicized on social media, the Inpaint feature provides pixel-perfect precision, giving users the ability to adjust items such as clothing, hair, and objects while preserving the Soul style.
The Inpaint operation is intuitive, similar to AI video editing found in other software. Users can select the components in the images they wish to alter, provide a prompt, and let the AI take care of the remainder. By “painting” over the elements to be modified, the AI interprets what to substitute according to the user’s description. Moreover, Inpaint permits users to upload actual photos for modification, not limited to those generated by Soul, thus enhancing its functionality. Users can generate an image using Soul and subsequently utilize Inpaint directly by masking the item and offering a prompt.
