{"id":1158232,"date":"2025-12-11T02:27:25","date_gmt":"2025-12-11T10:27:25","guid":{"rendered":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/?post_type=msr-blog-post&p=1158232"},"modified":"2025-12-11T02:27:27","modified_gmt":"2025-12-11T10:27:27","slug":"where-ai-meets-neuroscience-yansen-wangs-pursuit-of-human-centered-innovation","status":"publish","type":"msr-blog-post","link":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/articles\/where-ai-meets-neuroscience-yansen-wangs-pursuit-of-human-centered-innovation\/","title":{"rendered":"Where AI meets neuroscience: Yansen Wang\u2019s pursuit of human-centered innovation"},"content":{"rendered":"\n

\u201cCuriosity drives scientific breakthroughs, and the tools we create often reflect the human motivations behind that curiosity.\u201d<\/p>\n\n\n\n

For Yansen Wang<\/a>, a senior researcher at Microsoft Research Asia, this philosophy has guided his work at the intersection of AI and neuroscience.<\/p>\n\n\n\n

Wang\u2019s interest in science began early. While his classmates spent hours searching for information to complete an assignment, he solved it in 10 minutes by writing a few lines of code. Since then, programming became his way of tackling complex challenges, and the satisfaction of creating practical solutions fueled his passion for computer science. As his studies advanced, his goals crystallized: he wanted to create technology that genuinely serves people.<\/p>\n\n\n\n

\"Yansen
Yansen Wang, senior researcher at Microsoft Research Asia<\/figcaption><\/figure>\n\n\n\n

This people-centered approach has shaped Wang\u2019s career. As an undergraduate at Tsinghua University, he witnessed AI defeat the world\u2019s Go champion\u2014a moment that sparked a realization: AI could help us understand how humans think. This insight led him to pursue research in multimodal AI for his master\u2019s studies before joining Microsoft Research Asia \u2013 Shanghai. Today, in the AI\/ML Group, Wang works to bridge AI and neuroscience.<\/p>\n\n\n\n

A two-way journey between AI and neuroscience<\/h2>\n\n\n\n

Wang\u2019s research focuses on two complementary directions: advancing neuroscience by decoding how the brain works (AI for Brain), and drawing inspiration from the brain to improve AI architectures and algorithms (Brain for AI).<\/p>\n\n\n\n

For the AI for Brain project, Wang and his colleagues use noninvasive EEG(electroencephalogram) to build brain-computer interfaces, which record the electric signals from our brains without surgery, as their research platform. \u201cOur understanding of the brain is still very limited,\u201d he explains. \u201cIts multitasking abilities and rapid adaptability are far beyond what AI can achieve, so we\u2019re using AI to analyze the EEGsignals and uncover the links between perception, intention, and brain activity.\u201d<\/p>\n\n\n\n

The team has already made substantial progress. In terms of perception, they have decoded broad visual features, such as colors and simple moving scenes. Using diffusion models, they transformed neural signals into matching visual content and developed EEG2Video<\/a>, a baseline model that reconstructs video clips from EEG signals. To improve generalization across different contexts, the team has built multiple datasets linking EEG signals to everyday behaviors. See the video below for an example of this work.<\/p>\n\n\n\n

\"EEG2Video<\/figure>\n\n\n\n
\"EEG2Video
EEG2Video demo: The original video input is shown at the top and the reconstructed video at the bottom.<\/figcaption><\/figure>\n\n\n\n

For controlling with commands, Yansen and his team tackled the challenge of decoding letters from brain signals. They introduced an innovative codebook approach: instead of having participants imagine letter shapes, which are difficult for devices to recognize, they guided participants to imagine body movements, mathematical calculations, and other semantic information that devices could more easily recognize. AI then mapped these signals to specific letters. With portable devices, this method has achieved 30%\u201340% accuracy with 36 options (26 letters and 10 digits).<\/p>\n\n\n\n

\u201cWe\u2019re now working to extend this approach to controlling mobile phones and interacting on the web, exploring new interaction modes beyond letter input,\u201d he says.<\/p>\n\n\n\n

Still, the approach of using noninvasive brain-computer interfaces comes with many challenges. EEG signals have low signal-to-noise ratios and are easily disrupted by environmental and physiological interference, such as eye movements or muscle activity, making reliable readings difficult on portable devices. EEG data is scarce and must be collected in a controlled setting. And individual differences mean that systems often don\u2019t generalize well across users and tasks.<\/p>\n\n\n\n

\u201cWe\u2019re advancing research on EEG foundation models and hope to make them more robust with more data and larger models, much like large language models,\u201d Wang explains.<\/p>\n\n\n\n

Learning from the brain to make AI more efficient<\/h2>\n\n\n\n

For the Brain for AI project, Wang and his colleagues are exploring how brain function can address AI\u2019s energy demands.<\/p>\n\n\n\n

\u201cThe brain can accomplish tremendous thinking and computation with just the energy supplied from a bowl of rice, while AI requires vast resources and electricity to achieve similar results,\u201d he observes. \u201cEven more remarkable is that the brain efficiently handles many tasks without complex networks, like fine motor control. People only need a few examples to learn new tasks, but current AI models often need massive amounts of data to relearn. There must be design principles at play here that AI can learn from.\u201d<\/p>\n\n\n\n

The key differences lie in how neurons are structured and operate. Neurons use a spiking mechanism, firing and transmitting signals only when they reach their activation threshold. This results in extremely low energy consumption when they are at rest. Artificial neural networks, by contrast, perform large-scale computations even when there is very little information to process, using far more energy than the brain requires for similar tasks.<\/p>\n\n\n\n

Using this insight, Wang and his team developed a more efficient spiking neural network (SNN) framework<\/a>. In time-series prediction tasks, SNNs now perform comparably to traditional neural networks but can theoretically reduce energy consumption to a quarter of the latter, offering a new path for low-power AI (Figure 1).<\/p>\n\n\n\n

\"diagram\"
Figure 1. The SNN framework and workflow for time-series prediction<\/figcaption><\/figure>\n\n\n\n

\u201cThe spiking neural network research is just one part of our work,\u201d says Wang. \u201cNeurons in the brain are sparsely connected\u2014each one typically links to only a few nearby neurons, while in artificial neural networks, a single neuron connects to thousands of others. The brain\u2019s sparse connectivity also helps reduce energy consumption. If we continue learning from how the brain operates, AI will be able to generalize better and become more energy-efficient.\u201d<\/p>\n\n\n\n

\"Yansen
Yansen Wang gives a lecture at the TEDxBeijing Innovation Conference<\/figcaption><\/figure>\n\n\n\n

Breaking boundaries: From outsider to domain expert<\/h2>\n\n\n\n

For Wang, cross-disciplinary research in AI and neuroscience was new territory. He had no formal training in neuroscience or EEG analysis, but through dedicated study, active collaboration, and strong team support, he developed deep expertise.<\/p>\n\n\n\n

When he began EEG research, Wang studied medical textbooks and sought guidance from collaborating physicians. \u201cI set a rule: temporarily set aside my AI perspective and approach this as a doctor would. I studied medical textbooks thoroughly and learned to read EEG signals. Only then did I consider how AI could help.\u201d This approach helped him understand clinical problems rather than imposing familiar frameworks.<\/p>\n\n\n\n

Cross-disciplinary collaboration is not just about combining knowledge; it\u2019s about how different perspectives collide. For example, in research on epilepsy detection, Wang discovered that AI researchers and physicians approach problems differently. AI researchers often assume that with enough data, models can learn to identify features of epileptic seizures, such as spikes and slow waves. But physicians can quickly spot the rare abnormality in massive amount of hard-to-interpret EEG signals based on experience. Models can miss these abnormalities, even when trained on vast amounts of data.<\/p>\n\n\n\n

\u201cThis showed me that machine learning progress cannot rely solely on brute force. In data-scarce fields like medicine, we must incorporate domain expertise and build in the right assumptions to improve model performance.\u201d<\/p>\n\n\n\n

To help researchers build cross-domain knowledge, Microsoft Research Asia \u2013 Shanghai established a neuroscience study group, with weekly classes, homework, and discussions. After six months, Yansen had learned the fundamentals of neuroscience and gained practical guidance from senior researchers. \u201cThis collective learning atmosphere means we\u2019re no longer working in isolation but instead growing together as a community,\u201d he says.<\/p>\n\n\n\n

Microsoft Research Asia encourages open exploration and open exchange. At the Shanghai lab\u2019s weekly \u201cGrand Challenge\u201d meetings, researchers rigorously challenge one another\u2019s work. \u201cAt first, I wasn\u2019t used to this style of questioning,\u201d Wang admits. \u201cBut I realized that these challenges expose blind spots and allow research to improve through iterative refinement. The toughest questions often lead to the most important breakthroughs.\u201d<\/p>\n\n\n\n

\"a
Yansen Wang (third from right) discusses research questions with colleagues.<\/figcaption><\/figure>\n\n\n\n

Research with purpose: Building AI that serves people<\/h2>\n\n\n\n

For Wang, technology should serve people. Whether developing brain-computer interfaces or creating explainable AI for Go, the focus of the work should be on making AI useful and accessible.<\/p>\n\n\n\n

In 2022, Yansen and his colleagues launched what they called a \u201chuman salvation project\u201d for Go players. AI had surpassed top players, causing anxiety among professionals. Players could imitate AI moves but couldn\u2019t understand the reasoning behind them. They memorized patterns without developing their own strategic thinking. \u201cI thought, \u2018If AI could explain its logic, players could truly understand the strategies behind the moves,\u2019\u201d Wang says. \u201cWe wanted to help people improve alongside AI.\u201d Wang and the team are actively collaborating with Go lovers and professional Go players to verify the feasibility of the explanations. \u201cThat is so impressive!\u201d one of the teachers in a Go learning institute says, \u201cand I see the future of teaching human how to play Go\u201d.<\/p>\n\n\n\n

For Wang, this captures what drives his research: not the number of papers, but tangible impact. Perhaps it\u2019s the moment when a player grasps a brilliant move, or when someone finds more convenient ways to interact with devices, or when researchers apply new approaches for energy-efficient AI.<\/p>\n\n\n\n

\u201cAt Microsoft Research Asia, I can follow my interests and work with partners to solve meaningful problems for humanity,\u201d he says.<\/p>\n","protected":false},"excerpt":{"rendered":"

\u201cCuriosity drives scientific breakthroughs, and the tools we create often reflect the human motivations behind that curiosity.\u201d For Yansen Wang, a senior researcher at Microsoft Research Asia, this philosophy has guided his work at the intersection of AI and neuroscience. Wang\u2019s interest in science began early. While his classmates spent hours searching for information to […]<\/p>\n","protected":false},"author":34512,"featured_media":1150785,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr-content-parent":199560,"msr_hide_image_in_river":null,"footnotes":""},"research-area":[13556],"msr-locale":[268875],"msr-post-option":[],"class_list":["post-1158232","msr-blog-post","type-msr-blog-post","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-locale-en_us"],"msr_assoc_parent":{"id":199560,"type":"lab"},"_links":{"self":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1158232","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post"}],"about":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-blog-post"}],"author":[{"embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/users\/34512"}],"version-history":[{"count":5,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1158232\/revisions"}],"predecessor-version":[{"id":1158237,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-blog-post\/1158232\/revisions\/1158237"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/media\/1150785"}],"wp:attachment":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1158232"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1158232"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1158232"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=1158232"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}