人物经历

2010-2019年西北工业大学本科-博士[1]

2020年至今，中国人民大学高瓴人工智能学院，准聘助理教授

2019-2020年，百度研究院，人工智能研究员

主要成就

科研成果

1、DiHu,XuhongLi,LichaoMou,PuJin,DongChen,LipingJing,XiaoxiangZhu,andDejingDou,Cross-TaskTransferforMultimodalAerialSceneRecognition,InProceedingsoftheEuropeanConferenceonComputerVision(ECCV),2020.

2、RuiQian,DiHu,HeinrichDinkel,MengyueWu,NingXu,andWeiyaoLin,LearningtoVisuallyLocalizeMultipleSoundSourcesviaATwo-stageManner,InProceedingsoftheEuropeanConferenceonComputerVision(ECCV),2020.

3、DiHu,FeipingNie,andXuelongLi,DeepLinearDiscriminantAnalysisHashing,SciSinInform,2019.(CCFA)

4、DiHu,FeipingNie,andXuelongLi,DeepMultimodalClusteringforUnsupervisedAudiovisualLearning,InProceedingsoftheIEEEConferenceonComputerVisionandPatternRecognition(CVPR),2019.(CCFA)

5、DiHu,DongWang,FeipingNie,QiWang,andXuelongLi,ListentotheImage,InProceedingsoftheIEEEConferenceonComputerVisionandPatternRecognition(CVPR),2019.(CCFA)

6、DiHu,ChengzeWang,FeipingNie,andXuelongLi,DenseMultimodalFusionforHierarchicallyJointRepresentation,InProceedingsoftheIEEEConferenceonAcoustics,SpeechandSignalProcessing(ICASSP),2019.

7、DiHu,FeipingNie,andXuelongLi,DiscreteSpectralHashingforEfficientSimilarityRetrieval,IEEETrans.ImageProcessing(TIP),2018.(CCFA)

8、DiHu,FeipingNie,andXuelongLi,DeepBinaryReconstructionforCross-modalHashing,IEEETrans.Multimedia(TMM),2018.

9、XuelongLi,DiHu,andXiaoqiangLu,Image2song:SongRetrievalviaBridgingImageContentandLyricWords,InProceedingsoftheIEEEConferenceonComputerVision(ICCV),2017.(CCFA)

11、XuelongLi,DiHu,andFeipingNie,DeepBinaryReconstructionforCross-modalHashing,InProceedingsoftheACMConferenceonMultimedia(ACMMM),2017.(CCFA)

12、XuelongLi,DiHu,andFeipingNie,LargeGraphHashingwithSpectralRotation,InProceedingsoftheAAAIConferenceonArtificialIntelligence(AAAI),2017.(CCFA)

13、DiHu,XiaoqiangLu,andXuelongLi,MultimodalLearningviaExploringDeepSemanticSimilarity,InProceedingsoftheACMConferenceonMultimedia(ACMMM),2016.(CCFA)

14、DiHu,XuelongLi,andXiaoqiangLu,TemporalMultimodalLearninginAudiovisualSpeechRecognition,InProceedingsoftheIEEEConferenceonComputerVisionandPatternRecognition(CVPR),2016.(CCFA)

15、DiHu,ZhengWang,HaoyiXiong,DongWang,FeipingNie,andDejingDou,HeterogeneousSceneAnalysisviaSelf-supervisedAudiovisualLearning,InProceedingsoftheIEEEConferenceonComputerVisionandPatternRecognitionWorkshop(CVPRW),2020.

16、DiHu,LichaoMou,QingzhongWang,JunyuGao,YuanshengHua,DejingDou,andXiaoxiangZhu,DoesAmbientSoundHelp?-AudiovisualCrowdCounting,InProceedingsoftheIEEEConferenceonComputerVisionandPatternRecognitionWorkshop(CVPRW),2020.

17、YapengTian,DiHu,ChenliangXu,Co-LearnSoundingObjectVisualGroundingandVisuallyIndicatedSoundSeparationinACycle,InProceedingsoftheIEEEConferenceonComputerVisionandPatternRecognitionWorkshop(CVPRW),2020.

18、RuiQian,DiHu,HeinrichDinkel,MengyueWu,NingXu,WeiyaoLin,ATwo-StageFrameworkforMultipleSound-SourceLocalization,InProceedingsoftheIEEEConferenceonComputerVisionandPattern

获奖情况

2020.9荣获中国人工智能学会优秀博士论文奖

2019.8入选百度『AIDU』全球顶尖人工智能人才计划

2019.8荣获ACMXI’AN优秀博士论文奖（共2人）

2019.5入选CVPRDoctoralConsortium博士生论坛（大陆共4人）

2018.7荣获国家留学基金委赴卡内基梅隆大学联合培养学金

社会任职

期刊审稿人:TIP,TKDE,TMM,Neurocomputing

会议程序委员:NeurIPS2020-2021,CVPR201820202021,ICCV20192021,ECCV2020,ICML2021,AAAI201820202021,ACCV20182020

联合组织者:

CVPR2021TutorialonAudio-visualSceneUnderstanding

WACV2021TutorialonAudio-visualSceneUnderstanding

ICDM2019TutorialonAutomatedDeepLearning:Theory,Algorithms,Platforms,andApplications

学生要求

对客观存在保持好奇心，自驱，刻苦，以做有趣、有温度、有价值的研究为目标。

研究方向

机器多模态感知与学习：以大脑的多通道知觉为背景，挖掘并探究多模态信息（如图像、声音、触觉等）在机器感知、推理与理解等方向的潜在问题与方法，让机器具备『多感官认知能力』。