What if AI could interpret your imagination, turning images in your mind into reality? While that sounds like a detail in a science fiction, researchers have now accomplished exactly this. Researchers found that they could reconstruct highly accurate images from brain activity by using the popular stable diffusion (扩散) image generation model. The authors wrote that unlike previous studies, they didn’t need to train the AI models to create these images.

The researchers said that they first predicted a potential representation, which is a model of the image’s data, from fMRI (功能性磁共振成像) signals. Then, the model was processed and noise was added to it through the diffusion process. Finally, the researchers decoded (解码) text representations from fMRI signals and used them as input to produce a final constructed image.

The researchers wrote that a few studies have produced good reconstructions of images but it was only after training. This resulted in limitations because training complex models are challenging and there are not many samples to work with. Prior to this new study, no other researchers had tried using diffusion models for visual reconstruction.

In the past, we’ve seen other examples of how brainwaves and brain functions can create images. In 2014, a Shanghai - based artist Jody Xiong used sensors to connect sixteen people with disabilities to balloons of paint. The people would then use their thoughts to burst specific balloons and create their own paintings.

With the advancement of AI, more and more researchers have been testing the ways AI models can work with the human brain. In a January 2022 study, researchers at Radboud University in the Netherlands trained an AI network on fMRI data from 1, 050 unique faces and converted the brain imaging results into actual images. The study found that the AI was able to perform reconstruction. In the latest study, the researchers found that current diffusion models can now achieve high-quality visual reconstruction.

【小题1】How did researchers make images from brain activity in the previous research?

A．By using generation model.	B．By training AI models.
C．By reconstructing mind.	D．By predicting people’s imagination.

【小题2】Where are images’ data from?

A．The fMRI signals	B．The diffusion text.
C．The processed model.	D．The text representation.

【小题3】Why do the researchers think their new study is meaningful?

A．Their training work is very challenging.

B．They’ve made training models less complex.

C．Nobody has reconstructed images by brainwaves.

D．They can reconstruct accurate images from brain activity.

【小题4】What can we infer from the last paragraph?

A．Few researchers have used AI to make reconstruct images.

B．Scientists should do more to make high - quality reconstruction.

C．The AI models should be improved to work with fMRI data.

D．The AI contributes a lot to good visual reconstruction.