arXiv subjects

Wanying Ge

Publications and source records attributed to Wanying Ge.

21 records · Page 2Linked to original sources

arXiv2021

Raw Differentiable Architecture Search for Speech Deepfake and Spoofing Detection

End-to-end approaches to anti-spoofing, especially those which operate directly upon the raw signal, are starting to be competitive with their more traditional counterparts. Until recently, all such approaches consider only the learning of network parameters; the network architecture is still hand crafted. This too, however, can also be learned. Described in this paper is our attempt to learn automatically the network architecture of a speech deepfake and spoofing detection solution, while jointly optimising other network components and parameters, such as the first convolutional layer which operates on raw signal inputs. The resulting raw differentiable architecture search system delivers a tandem detection cost function score of 0.0517 for the ASVspoof 2019 logical access database, a result which is among the best single-system results reported to date.

eess.AS↗

arXiv2021

Partially-Connected Differentiable Architecture Search for Deepfake and Spoofing Detection

This paper reports the first successful application of a differentiable architecture search (DARTS) approach to the deepfake and spoofing detection problems. An example of neural architecture search, DARTS operates upon a continuous, differentiable search space which enables both the architecture and parameters to be optimised via gradient descent. Solutions based on partially-connected DARTS use random channel masking in the search space to reduce GPU time and automatically learn and optimise complex neural architectures composed of convolutional operations and residual blocks. Despite being learned quickly with little human effort, the resulting networks are competitive with the best performing systems reported in the literature. Some are also far less complex, containing 85% fewer parameters than a Res2Net competitor.

cs.LG↗

arXiv2019

Enhancing Interconnect Reliability and Performance by Converting Tantalum to 2D Layered Tantalum Sulfide at Low Temperature

The interconnect half-pitch size will reach ~20 nm in the coming sub-5 nm technology node. Meanwhile, the TaN/Ta (barrier/liner) bilayer stack has to be > 4 nm to ensure acceptable liner and diffusion barrier properties. Since TaN/Ta occupy a significant portion of the interconnect cross-section and they are much more resistive than Cu, the effective conductance of an ultra-scaled interconnect will be compromised by the thick bilayer. Therefore, two dimensional (2D) layered materials have been explored as diffusion barrier alternatives. However, many of the proposed 2D barriers are prepared at too high temperatures to be compatible with the back-end-of-line (BEOL) technology. In addition, as important as the diffusion barrier properties, the liner properties of 2D materials must be evaluated, which has not yet been pursued. Here, a 2D layered tantalum sulfide (TaSx) with ~1.5 nm thickness is developed to replace the conventional TaN/Ta bilayer. The TaSx ultra-thin film is industry-friendly, BEOL-compatible, and can be directly prepared on dielectrics. Our results show superior barrier/liner properties of TaSx compared to the TaN/Ta bilayer. This single-stack material, serving as both a liner and a barrier, will enable continued scaling of interconnects beyond 5 nm node.

cond-mat.mtrl-sci↗