Japanese scientists announce the results of research on artificial intelligence: let AI learn to "decode"

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 becoming increasingly real. At the end of December 2017, four scientists from the Shengu Research Laboratory at Kyoto University in Japan announced their work on artificial intelligence (AI). Although still in its early stages, the team has developed new technologies that allow AI to "decode" brainwave activity. The AI system created by the team can interpret brainwave patterns and generate corresponding images based on what a person is seeing or thinking about. This technology, which visualizes the perception process through brain activity, is called "deep image reconstruction." Once brain activity is scanned, the computer can reverse-engineer the data, producing a visual representation of how the AI interprets and reconstructs the information. Professor Yukiyasu Kamitani from Kyoto University told CNBC that the team has been working on reconstructing and reproducing the images people see by observing brain activity. In previous studies, they focused on simple shapes or pixels, but now they have a better understanding of how the human brain processes visual information. He explained, "Now we know that the brain processes visual information by extracting different levels of features and components of varying complexity." By understanding this, the neural network used in the study can mimic the brain's hierarchical structure, allowing AI to detect and understand objects more accurately than before. It no longer just recognizes binary pixels but can represent complex visual elements. In the deep image reconstruction algorithm, AI starts with a random image and gradually adjusts pixel values based on the features detected in the brainwave data. The team also used a Deep Generator Network (DGN) to ensure the generated images look natural. During the study, the researchers tested various stimuli, including natural scenes, artificial geometry, and letters. In some cases, participants viewed images while their brainwaves were recorded, and in others, they recalled images after viewing them. Unlike previous methods, the AI can attempt to reconstruct images even when a person is recalling a memory, though this process takes longer due to less intense brain activation. Through continuous learning from brainwave data, the AI improves its ability to interpret mental imagery. While it doesn’t truly "see" thoughts, it can recognize the brainwave patterns associated with specific mental states. Although the current results are still limited, the potential applications of this technology could be revolutionary as accuracy improves. Imagine being able to create art simply by imagining it, or even visualizing dreams through a computer. For patients with certain mental conditions, this could provide valuable external support. Professor Kotani Kang believes that, with advancements in AI and brain-computer interfaces, one day, communication through thought and imagination may become a reality.

Oil Immersed Transformer

Oil Immersed Transformer,Hermetically Sealed Oil Filled Transformer,Immersed Transformer,Oil Filled Distribution Transformers

Tianhong Electric Power Technology Co., Ltd , https://www.tianhongtransformer.com