3 Hits in 1.8 sec

Understanding the Tradeoffs in Client-side Privacy for Downstream Speech Tasks [article]

Peter Wu, Paul Pu Liang, Jiatong Shi, Ruslan Salakhutdinov, Shinji Watanabe, Louis-Philippe Morency
2021 arXiv   pre-print
As users increasingly rely on cloud-based computing services, it is important to ensure that uploaded speech data remains private.  ...  These insights call for more research in client-side privacy to ensure a safer deployment of cloud-based speech processing services.  ...  Based on these insights, we propose several extensions for future work and call for more research in client-side privacy to ensure safe cloud-based speech processing.  ... 
arXiv:2101.08919v2 fatcat:ufeaku4zlvfv5mn6fivxuh22ua

A Review of 1D Convolutional Neural Networks toward Unknown Substance Identification in Portable Raman Spectrometer [article]

M. Hamed Mozaffari, Li-Lin Tay
2020 arXiv   pre-print
Different hyperparameters, such as the number of convolutional layers and kernel sizes, are assessed using two 1D CNN structures by (Hu et al., 2019) .  ...  methods (e.g., Rolling ball, Rubberband) (Kneen & Annegarn, 1996) Least Squares Curve Fitting (Baek et al., 2015; Lieber & Mahadevan-Jansen, 2003; Z.  ... 
arXiv:2006.10575v1 fatcat:rlaraaztznhrjp3xw7qn4a56sm

Impact Of Semantics, Physics And Adversarial Mechanisms In Deep Learning [article]

Ilya Kavalerov
To study this question, we develop a dataset of mixtures containing arbitrary sounds, and use it to investigate the space of mask-based separation architectures, varying both the overall network architecture  ...  −60, f source +60]) Hz using the high quality C++ sound library Rubberband [115].  ...  I would also like to acknowledge support of Tensorflow Research Cloud (TFRC) for their generosity in making available compute time on TPUs for me, and the Google Cloud Platform (GCP) Research Credits for  ... 
doi:10.13016/bq0r-adgr fatcat:sg327rrqpnevjht4yu5m5dncfm