We present a large-scale study exploring the capability of temporal deep neural networks to interpret natural human kinematics and introduce the first method for active biometric authentication with mobile inertial sensors. At Google, we have created a first-of-its-kind data set of human movements, passively collected by 1500 volunteers using their smartphones daily over several months. We compare several neural architectures for efficient learning of temporal multi-modal data representations, propose an optimized shift-invariant dense convolutional mechanism, and incorporate the discriminatively trained dynamic features in a probabilistic generative framework taking into account temporal characteristics. Our results demonstrate that human kinematics convey important information about user identity and can serve as a valuable component of multi-modal authentication systems. Finally, we demonstrate that the proposed model can also be successfully applied in a visual context.

Abstract-Convolutional Neural Networks (CNNs) have gained significant traction in the field of machine learning, particularly due to their high accuracy in visual recognition. Recent works have pushed the performance of GPU implementations of CNNs to significantly improve their classification and training times. With these improvements, many frameworks have become available for implementing CNNs on both CPUs and GPUs, with no support for FPGA implementations. In this work we present a modified version of the popular CNN framework Caffe, with FPGA support. This allows for classification using CNN models and specialized FPGA implementations with the flexibility of reprogramming the device when necessary, seamless memory transactions between host and device, simple-to-use test benches, and the ability to create pipelined layer implementations. To validate the framework, we use the Xilinx SDAccel environment to implement an FPGA-based Winograd convolution engine and show that the FPGA layer can be used alongside other layers running on a host processor to run several popular CNNs (AlexNet, GoogleNet, VGG A, Overfeat). The results show that our framework achieves 50 GFLOPS across 3 × 3 convolutions in the benchmarks. This is achieved within a practical framework, which will aid in future development of FPGA-based CNNs.

Abstract-Advances in diffuse reflectance infra-red spectroscopy measurements have made it possible to estimate a number of functional properties of soil inexpensively and accurately. Core to such techniques are machine learning methods that can map high-dimensional spectra to real-valued outputs. While previous works have considered predicting each property individually using simple regression methods, the correlation structure present in the output variables prompts us to consider methods that can leverage this structure to make more accurate predictions. In this paper, we leverage advances in deep learning architectures, specifically convolutional neural networks and conditional restricted Boltzmann machines for structured output prediction for soil property prediction. We evaluate our methods on two recent spectral datasets, where output soil properties are shown to have a measurable degree of correlation.

No abstract

No abstract