Pre-polymerized vinyl trimethoxy silane(PVTMS)@MWCNT nano-aerogel system was constructed via radical polymerization,sol-gel transition and supercritical CO_(2)drying.The fabricated organic-inorganic hybrid PVTMS@MWCNT...Pre-polymerized vinyl trimethoxy silane(PVTMS)@MWCNT nano-aerogel system was constructed via radical polymerization,sol-gel transition and supercritical CO_(2)drying.The fabricated organic-inorganic hybrid PVTMS@MWCNT aerogel structure shows nano-pore size(30-40 nm),high specific surface area(559 m^(2)g^(−1)),high void fraction(91.7%)and enhanced mechanical property:(1)the nano-pore size is beneficial for efficiently blocking thermal conduction and thermal convection via Knudsen effect(beneficial for infrared(IR)stealth);(2)the heterogeneous interface was beneficial for IR reflection(beneficial for IR stealth)and MWCNT polarization loss(beneficial for electromagnetic wave(EMW)attenuation);(3)the high void fraction was beneficial for enhancing thermal insulation(beneficial for IR stealth)and EMW impedance match(beneficial for EMW attenuation).Guided by the above theoretical design strategy,PVTMS@MWCNT nano-aerogel shows superior EMW absorption property(cover all Ku-band)and thermal IR stealth property(ΔT reached 60.7℃).Followed by a facial combination of the above nano-aerogel with graphene film of high electrical conductivity,an extremely high electromagnetic interference shielding material(66.5 dB,2.06 mm thickness)with superior absorption performance of an average absorption-to-reflection(A/R)coefficient ratio of 25.4 and a low reflection bandwidth of 4.1 GHz(A/R ratio more than 10)was experimentally obtained in this work.展开更多
Recently,segmentation-based scene text detection has drawn a wide research interest due to its flexibility in describing scene text instance of arbitrary shapes such as curved texts.However,existing methods usually ne...Recently,segmentation-based scene text detection has drawn a wide research interest due to its flexibility in describing scene text instance of arbitrary shapes such as curved texts.However,existing methods usually need complex post-processing stages to process ambiguous labels,i.e.,the labels of the pixels near the text boundary,which may belong to the text or background.In this paper,we present a framework for segmentation-based scene text detection by learning from ambiguous labels.We use the label distribution learning method to process the label ambiguity of text annotation,which achieves a good performance without using additional post-processing stage.Experiments on benchmark datasets demonstrate that our method produces better results than state-of-the-art methods for segmentation-based scene text detection.展开更多
In this paper,we develop an orthogonal frequency-division multiplexing(OFDM)-based over-theair(OTA)aggregation solution for wireless federated learning(FL).In particular,the local gradients in massive Internet of thin...In this paper,we develop an orthogonal frequency-division multiplexing(OFDM)-based over-theair(OTA)aggregation solution for wireless federated learning(FL).In particular,the local gradients in massive Internet of things(IoT)devices are modulated by an analog waveform and are then transmitted using the same wireless resources.To this end,achieving perfect waveform superposition is the key challenge,which is difficult due to the existence of frame timing offset(TO)and carrier frequency offset(CFO).In order to address these issues,we propose a two-stage waveform pre-equalization technique with a customized multiple access protocol that can estimate and then mitigate the TO and CFO for the OTA aggregation.Based on the proposed solution,we develop a hardware transceiver and application software to train a real-world FL task,which learns a deep neural network to predict the received signal strength with the global positioning system information.Experiments verify that the proposed OTA aggregation solution can achieve comparable performance to offline learning procedures with high prediction accuracy.展开更多
基金the National Natural Science Foundation(No.52073187)NSAF Foundation(No.U2230202)for their financial support of this project+3 种基金National Natural Science Foundation(No.51721091)Programme of Introducing Talents of Discipline to Universities(No.B13040)State Key Laboratory of Polymer Materials Engineering(No.sklpme2022-2-03)support of China Scholarship Council
文摘Pre-polymerized vinyl trimethoxy silane(PVTMS)@MWCNT nano-aerogel system was constructed via radical polymerization,sol-gel transition and supercritical CO_(2)drying.The fabricated organic-inorganic hybrid PVTMS@MWCNT aerogel structure shows nano-pore size(30-40 nm),high specific surface area(559 m^(2)g^(−1)),high void fraction(91.7%)and enhanced mechanical property:(1)the nano-pore size is beneficial for efficiently blocking thermal conduction and thermal convection via Knudsen effect(beneficial for infrared(IR)stealth);(2)the heterogeneous interface was beneficial for IR reflection(beneficial for IR stealth)and MWCNT polarization loss(beneficial for electromagnetic wave(EMW)attenuation);(3)the high void fraction was beneficial for enhancing thermal insulation(beneficial for IR stealth)and EMW impedance match(beneficial for EMW attenuation).Guided by the above theoretical design strategy,PVTMS@MWCNT nano-aerogel shows superior EMW absorption property(cover all Ku-band)and thermal IR stealth property(ΔT reached 60.7℃).Followed by a facial combination of the above nano-aerogel with graphene film of high electrical conductivity,an extremely high electromagnetic interference shielding material(66.5 dB,2.06 mm thickness)with superior absorption performance of an average absorption-to-reflection(A/R)coefficient ratio of 25.4 and a low reflection bandwidth of 4.1 GHz(A/R ratio more than 10)was experimentally obtained in this work.
基金supported by the National Key R&D Program of China(2018AAA0100104,2018AAA0100100)the National Natural Science Foundation of China(Grant No.61702095)the Natural Science Foundation of Jiangsu Province(BK20211164).
文摘Recently,segmentation-based scene text detection has drawn a wide research interest due to its flexibility in describing scene text instance of arbitrary shapes such as curved texts.However,existing methods usually need complex post-processing stages to process ambiguous labels,i.e.,the labels of the pixels near the text boundary,which may belong to the text or background.In this paper,we present a framework for segmentation-based scene text detection by learning from ambiguous labels.We use the label distribution learning method to process the label ambiguity of text annotation,which achieves a good performance without using additional post-processing stage.Experiments on benchmark datasets demonstrate that our method produces better results than state-of-the-art methods for segmentation-based scene text detection.
基金This work was supported by Innovation and Technology Fund under Grant GHP/016/18GD and Guangdong Special Fund for Science and Technology Development under Grant 2019A050503001.
文摘In this paper,we develop an orthogonal frequency-division multiplexing(OFDM)-based over-theair(OTA)aggregation solution for wireless federated learning(FL).In particular,the local gradients in massive Internet of things(IoT)devices are modulated by an analog waveform and are then transmitted using the same wireless resources.To this end,achieving perfect waveform superposition is the key challenge,which is difficult due to the existence of frame timing offset(TO)and carrier frequency offset(CFO).In order to address these issues,we propose a two-stage waveform pre-equalization technique with a customized multiple access protocol that can estimate and then mitigate the TO and CFO for the OTA aggregation.Based on the proposed solution,we develop a hardware transceiver and application software to train a real-world FL task,which learns a deep neural network to predict the received signal strength with the global positioning system information.Experiments verify that the proposed OTA aggregation solution can achieve comparable performance to offline learning procedures with high prediction accuracy.