Posts by Collection

publications

Effect of deposition potential and concentration of electroactive substances on Bi2Te3 nanowires fabricated by electrochemical method

Published in Nanotechnology, 2019

In this paper, Bi2Te3 nanowires were prepared in anodized aluminum oxide template by electrochemical deposition. The morphological microstructural and electrical resistance characteristics of the nanowires were discussed to reveal the effect of deposition potential and electroactive substance (HTeO2 +) concentration. According to the electrode dynamics formula, it is found that the increase of electrode potential leads to the decrease of deposition current, so that deposition rate of nanowires decreases. At the same time, the deposition current controlled by diffusion in the mass transport process will have a maximum value with the increasing of deposition time. The deposition potential determines the favorable crystal plane for nanowires growth by the selection of proper surface energy. The temperature dependence of resistances in Bi2Te3 nanowires fabricated under different concentration of HTeO2 + reveals the transformation of the carriers’ main scattering mechanism. This study could provide a better understanding of the deposition process of Bi2Te3 nanowires.

Recommended citation: Wenxin L, Wangchen Z, Yanpeng Z, et al. Effect of deposition potential and concentration of electroactive substances on Bi2Te3 nanowires fabricated by electrochemical method[J]. Nanotechnology, 2019, 30(24): 245702.

Flexible radiative cooling material based on amorphous alumina nanotubes

Published in Optical Materials Express, 2020

With the rapid development in near / far field thermal radiation and micro- / nano- fabrication, passive radiative cooling has become an intriguing topic in both fundamental scientific research and practical energy engineering. In this paper, we use Amorphous Alumina Nanotubes (AANs) to prepare a flexible material for high-efficient daytime radiative cooling. Instead of applying parallel nanotube array or total randomly distributed nanotubes, we experimentally fabricated a porous membrane by introducing hexagonal lattice roots at the bottom and random agglomeration at the top for AANs. Near-unity emissivity originated from alumina absorption and complex scattering inside the membrane covers the 8-13 µm atmosphere window. Under direct sunlight, the flexible AANs membrane achieves a theoretical net cooling power of 71.0 W/m2, leading to an experimental maximum temperature reduction of 6.7 °C to the ambient air. Our material paves herein a way for producing low-cost and efficient flexible daytime radiative coolers.

Recommended citation: Zhou Y, Liu Y, Li Y, et al. Flexible radiative cooling material based on amorphous alumina nanotubes[J]. Optical Materials Express, 2020, 10(7): 1641-1648.

AI as an Active Writer: Interaction strategies with generated text in human-AI collaborative fiction writing

Published in Joint Proceedings of the ACM IUI Workshops, 2022

Machine Learning (ML) has become an important part of the creative process for human fiction writers, allowing them to utilize various sources of information and be inspired by strategies and data previously seldom explored. To investigate how human writers collaborate with ML systems in fiction writing, we prototyped a web-based human-AI collaborative writing tool that allows writers to shorten, edit, summarize, and regenerate text produced by AI. To investigate the dynamics of human-AI interaction in fiction co-writing, we used a “finish each other’s story” approach where humans and machines took turns writing collaborative fiction. In results from a preliminary study with 9 users, we found that users took inspiration from unexpected text generated by the machine, that users expected reduced fluency and coherence in the machine text when allowed to edit the output, and that they perceived a mental model of the AI as an active writer in the collaborative process rather than simply as a passive AI writing assistant. This study provides design implications on supporting co-creative writing of humans and machines.

Recommended citation: Yang D, Zhou Y, Zhang Z, et al. AI as an Active Writer: Interaction strategies with generated text in human-AI collaborative fiction writing[C]//Joint Proceedings of the ACM IUI Workshops. CEUR-WS Team, 2022, 10: 1-11.
Download Paper

Investigating robustness of biological vs. backprop based learning

Published in ICASSP, 2022

Robustness of learning algorithms remains an important prob- lem to be solved from both the perspective of adversarial at- tacks and improving generalization. In this work, we inves- tigate the robustness of biologically inspired Hebbian learn- ing algorithm in depth. We find that Hebbian learning based algorithms outperform conventional learning algorithms like CNNs by a huge margin of upto 18% on the CIFAR-10 dataset under the addition of noise. We highlight that an important reason for this is the underlying representations that are being learnt by the learning algorithms. Specifically, we find that the Hebbian method learns the most robust representations compared to other methods that helps it to generalize better. We also conduct ablations on the Hebbian network and show- case that robustness of the model drops by upto 16% on the CIFAR-10 dataset if the representation capacity of the net- work is deteriorated. Hence, we find that the representations learnt play an important role in the resultant robustness of the models. We conduct experiments on multiple datasets and show that the results hold on all the datasets and at various noise levels.

Recommended citation: Zhou Y, Wang M, Gupta M, et al. Investigating robustness of biological vs. backprop based learning[C]//ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2022: 3533-3537.

Zhou Yanpeng(周滟朋)