Imagine a world where a computer can read your thoughts or what you're looking at—sounds like something out of a science fiction movie. However, thanks to the groundbreaking research conducted by a Japanese team, this futuristic concept is slowly becoming a reality.
At the end of December 2017, four scientists from the Shengu Research Laboratory at Kyoto University announced their progress in artificial intelligence (AI). Although still in its early stages, the team has developed innovative technologies that allow AI to "decode" brainwave patterns.
The AI created by the team can interpret brain activity and generate corresponding, albeit blurry, images based on what a person is seeing or thinking. This technology, which visualizes the perception process from brain activity, is called "deep image reconstruction."
Once brain fluctuations are scanned, the computer decodes the signals and produces a visual representation of how the AI interprets and reconstructs the information.
Professor Yukiyasu Kamitani explained to CNBC that the team has been working on reconstructing and reproducing images people see by observing brain activity. Previously, they assumed images were made up of simple shapes or pixels, but now they have a better understanding of how the human brain processes visual information.
He said, "Now we know that the brain processes visual information by extracting features at different levels of complexity." With this knowledge, the AI model used in the study can represent the brain's hierarchical structure, allowing it to better detect and understand objects seen by the brain—not just as binary pixels, but in more detailed forms.
In the deep image reconstruction algorithm, the AI starts with a random image and gradually adjusts pixel values based on the decoded brainwave features. The team also used a Deep Generator Network (DGN) to ensure the generated images look natural.
During the study, the team tested natural scenes, geometric shapes, and letters. In some cases, participants viewed images while their brain activity was recorded. In others, they recalled images, and the AI attempted to reconstruct them.
Unlike previous methods, even when recalling an image, the AI could attempt to create a visual representation. However, due to less brain activation during recall, the analysis took longer.
Through continuous learning from brainwave data, the AI improves its ability to interpret thoughts. Strictly speaking, the AI doesn't truly "see" thoughts, but it learns what brainwaves look like when a person thinks about certain things.
Although the current results are limited, the potential applications of this technology are vast. As accuracy improves, it could revolutionize communication, art creation, and even help those with mental illnesses by visualizing internal experiences.
Professor Kotani Kang believes that with further advancements in AI and brain-computer interfaces, one day, communication through thought and imagination may become possible.
10 KV Oil Immersed Transformer
10 Kv Oil Immersed Transformer,10Kv Oil Immersed Transformer,500Kva/630Kva 10Kv Oil Immersed Transformer,Efficient 10 Kv Oil Immersed Transformer
Tianhong Electric Power Technology Co., Ltd , https://www.tianhongtransformer.com