A membrane bioreactor(MBR)with ordinary aeration(reactor R1)and a MBR with microporous aeration(reactor R2)are conducted in parallel to investigate the characteristics of oily sludge and its effect on membrane fouling...A membrane bioreactor(MBR)with ordinary aeration(reactor R1)and a MBR with microporous aeration(reactor R2)are conducted in parallel to investigate the characteristics of oily sludge and its effect on membrane fouling.The results indicate that the order of membrane fouling rate from high to low is:reactor R1 dissolved oxygen(DO)(1 mg/L)>reactor R2 DO(1 mg/L)>reactor R1 DO(4 mg/L)>reactor R2 DO(4 mg/L).Membrane fouling rate is not related to oily sludge concentration but to oily sludge sedimentation performance and the small particle oily sludge is the key factor to affect the membrane fouling.The soluble microbial products(SMP)are examined by three-dimensional excitation-emission matrix(3DEEM)fluorescence spectra.3DEEM spectra demonstrate that the main organic substances of the SMP in two reactors are tyrosine aromatic protein,fulvic acid-like substances and soluble microbial products under DO are1 mg/L and 4 mg/L,respectively.The proportion sum of fulvic acid-like substances and soluble microbial products is the key factor affecting membrane fouling,and membrane fouling accelerates as the ratio increases.展开更多
Text-to-video retrieval(TVR)has made significant progress with advances in vision and language representation learning.Most existing methods use real-valued and hash-based embeddings to represent the video and text,al...Text-to-video retrieval(TVR)has made significant progress with advances in vision and language representation learning.Most existing methods use real-valued and hash-based embeddings to represent the video and text,allowing retrieval by computing their similarities.However,these methods are often inefficient for large volumes of video,and require significant storage and computing resources.In this work,we present a plug-and-play multi-modal multi-tagger-driven pre-screening framework,which pre-screens a substantial number of videos before applying any TVR algorithms,thereby efficiently reducing the search space of videos.We predict discrete semantic tags for video and text with our proposed multi-modal multi-tagger module,and then leverage an inverted index for space-efficient and fast tag matching to filter out irrelevant videos.To avoid filtering out relevant videos for text queries due to inconsistent tags,we utilize contrastive learning to align video and text embeddings,which are then fed into a shared multi-tag head.Extensive experimental results demonstrate that our proposed method significantly accelerates the TVR process while maintaining high retrieval accuracy on various TVR datasets.展开更多
基金National Natural Science Foundation of China(No.21173026)Key Program of the Natural Science Foundation of Hubei Province(No.2013CFA107).
文摘A membrane bioreactor(MBR)with ordinary aeration(reactor R1)and a MBR with microporous aeration(reactor R2)are conducted in parallel to investigate the characteristics of oily sludge and its effect on membrane fouling.The results indicate that the order of membrane fouling rate from high to low is:reactor R1 dissolved oxygen(DO)(1 mg/L)>reactor R2 DO(1 mg/L)>reactor R1 DO(4 mg/L)>reactor R2 DO(4 mg/L).Membrane fouling rate is not related to oily sludge concentration but to oily sludge sedimentation performance and the small particle oily sludge is the key factor to affect the membrane fouling.The soluble microbial products(SMP)are examined by three-dimensional excitation-emission matrix(3DEEM)fluorescence spectra.3DEEM spectra demonstrate that the main organic substances of the SMP in two reactors are tyrosine aromatic protein,fulvic acid-like substances and soluble microbial products under DO are1 mg/L and 4 mg/L,respectively.The proportion sum of fulvic acid-like substances and soluble microbial products is the key factor affecting membrane fouling,and membrane fouling accelerates as the ratio increases.
基金supported by the National Natural Science Foundation of China(No.62476188)the Open Projects Program of the State Key Laboratory of Multimodal Artificial Intelligence Systems,and the Academy of Finland in the USSEE Project(No.345791).
文摘Text-to-video retrieval(TVR)has made significant progress with advances in vision and language representation learning.Most existing methods use real-valued and hash-based embeddings to represent the video and text,allowing retrieval by computing their similarities.However,these methods are often inefficient for large volumes of video,and require significant storage and computing resources.In this work,we present a plug-and-play multi-modal multi-tagger-driven pre-screening framework,which pre-screens a substantial number of videos before applying any TVR algorithms,thereby efficiently reducing the search space of videos.We predict discrete semantic tags for video and text with our proposed multi-modal multi-tagger module,and then leverage an inverted index for space-efficient and fast tag matching to filter out irrelevant videos.To avoid filtering out relevant videos for text queries due to inconsistent tags,we utilize contrastive learning to align video and text embeddings,which are then fed into a shared multi-tag head.Extensive experimental results demonstrate that our proposed method significantly accelerates the TVR process while maintaining high retrieval accuracy on various TVR datasets.
