Results 11 to 20 of about 738,909 (270)
Net-Shape NiTi Shape Memory Alloy by Spark Plasma Sintering Method
Applied Sciences, 2021 An analysis of the shape memory effect of a NiTi alloy by using the spark plasma sintering approach has been carried out. Spark plasma sintering of Ti50Ni50 powder (20–63 µm) at a temperature of 900 °C produced specimens showing good shape memory effects.
Sneha Samal +8 more
doaj +1 more source
BM-NAS: Bilevel Multimodal Neural Architecture Search
Proceedings of the AAAI Conference on Artificial Intelligence, 2022 Deep neural networks (DNNs) have shown superior performances on various multimodal learning problems. However, it often requires huge efforts to adapt DNNs to individual multimodal tasks by manually engineering unimodal features and designing multimodal feature fusion strategies. This paper proposes Bilevel Multimodal Neural Architecture Search (BM-NAS)
Yin, Yihang, Huang, Siyu, Zhang, Xiang
openaire +2 more sources
BN-NAS: Neural Architecture Search with Batch Normalization [PDF]
2021 IEEE/CVF International Conference on Computer Vision (ICCV), 2021 We present BN-NAS, neural architecture search with Batch Normalization (BN-NAS), to accelerate neural architecture search (NAS). BN-NAS can significantly reduce the time required by model training and evaluation in NAS. Specifically, for fast evaluation, we propose a BN-based indicator for predicting subnet performance at a very early training stage ...
Chen, Boyu +7 more
openaire +2 more sources
Wrinkling prediction, formation and evolution in thin films adhering on polymeric substrata
Materials & Design, 2020 Wrinkling has recently attracted an increasing interest by suggesting a number of unforeseeable applications in many emerging material science and engineering fields.
A. Cutolo +5 more
doaj +1 more source
Rethinking Architecture Selection in Differentiable NAS
, 2021 Differentiable Neural Architecture Search is one of the most popular Neural Architecture Search (NAS) methods for its search efficiency and simplicity, accomplished by jointly optimizing the model weight and architecture parameters in a weight-sharing supernet via gradient-based algorithms.
Wang, Ruochen +4 more
openaire +2 more sources
M-NAS: Meta Neural Architecture Search
Proceedings of the AAAI Conference on Artificial Intelligence, 2020 Neural Architecture Search (NAS) has recently outperformed hand-crafted networks in various areas. However, most prevalent NAS methods only focus on a pre-defined task. For a previously unseen task, the architecture is either searched from scratch, which is inefficient, or transferred from the one obtained on some other task, which might be sub-optimal.
Jiaxing Wang +3 more
openaire +2 more sources
Applied Sciences, 2020 The aim of this article is to investigate the interfacial adhesion of Ni-Ti shape memory alloy with a polymer matrix of Poly (methyl methacrylate) (PMMA). The surface pattern on Ni-Ti plates was channeled by a solid state laser machine. The laser machine
Sneha Samal +7 more
doaj +1 more source
Applied Sciences, 2021 (1) Methods for checking the condition of monumental masonry structures can still be considered understudied. Among the different approaches available in the literature, the graphical ones have a special role, due to their simplicity and effectiveness ...
Concetta Cusano +4 more
doaj +1 more source
NAS-FCOS: Efficient Search for Object Detection Architectures [PDF]
International Journal of Computer Vision, 2021 14 pages, 11 figures, journal extension of NAS-FCOS (CVPR 2020), accepted by IJCV.
Ning Wang +6 more
openaire +2 more sources
Journal of Agricultural Engineering, 2017 The long-standing awareness of the environmental impact of land-use change (LUC) has led scientific community to develop tools able to predict their amount and to evaluate their effect on environment, with the aim supporting policy makers in their ...
Stefania Pindozzi +4 more
doaj +1 more source