報告題目：Towards the next generation of AI: visual computation with neural spikes
報 告 人：劉健教授 英國萊斯特大學
報告時間：2020年6月4日 16:00-17:00
報告地點：騰訊會議ID：292 189 248
或點擊鍊接直接加入會議：
https://meeting.tencent.com/s/WOYmez8MFYyi
校内聯系人：孫維鵬 sunwp@jlu.edu.cn
報告摘要：
Neuromorphic computing has been suggested as the next generation of computational strategy. In neuroscience, neural coding is for understanding how the brain processes stimulus from the environment, moreover, it is also a cornerstone for designing algorithms of brain-machine inference. Here, I will show some of the recent progress that has been achieved in data-driven visual computation models that use neural spikes to analyze natural scenes. I hypothesize that we need a hyper-circuit view of neural network computing framework, in which specific computations are utilized by different network motifs inspired by techniques of probabilistic graph models and deep learning. As a proof of concept, the revealed mechanisms and proposed algorithms can provide new insights into neuromorphic computing for next-generation of general-purpose AI.
報告人簡介：
劉健博士，吉林大學數學學士，北京大學力學碩士，美國加州大學洛杉矶分校數學博士，先後為法國國家科學院和哥廷根大學博士後研究員。曾為奧地利格拉茲理工大學理論計算機研究所助理教授，參與歐盟旗艦研究項目“人腦計劃”。現為英國萊斯特大學系統神經科學中心長聘助理教授。研究領域為計算神經科學與類腦智能，神經網絡學習與記憶，視覺系統編碼。近年來在Nature communications, eLife, Journal of neuroscience, PLoS computational biology, IEEE Transactions on Neural Networks and Learning Systems, IEEE Transactions on cybernetics等頂級期刊發表多篇論文。研究資助方包括歐盟研究委員會，英國皇家學會牛頓高級學者基金，中國之江實驗室。
