November 14, 2016

Sparsification and Separation of Deep Learning Layers for Constrained Resource Inference on Wearables

  • Bhattacharya S.
  • Lane N.

Deep learning has revolutionized the way sensor data are analyzed and interpreted. The accuracy gains these approaches offer make them attractive for the next generation of mobile, wearable and embedded sensory applications. However, state-of-the-art deep learning algorithms typically require a significant amount of device and processor resources, even just for the inference stages that are used to discriminate high-level classes from low-level data. The limited availability of memory, computation, and energy on mobile and embedded platforms thus pose a significant challenge to the adoption of these powerful learning techniques. In this paper, we propose SparseSep, a new approach that leverages the sparsification of fully connected layers and separation of convolutional kernels to reduce the resource requirements of popular deep learning algorithms. As a result, SparseSep allows large-scale DNNs and CNNs to run e ciently on mobile and embedded hardware with only minimal impact on inference accuracy. We experiment using SparseSep across a variety of common processors such as the Qualcomm Snapdragon 400, ARM Cortex M0 and M3, and Nvidia Tegra K1, and show that it allows inference for various deep models to execute more e ciently; for example, on average requiring 11.3 times less memory and running 13.3 times faster on these representative platforms.

View Original Article

Recent Publications

January 01, 2018

Fair Dynamic Spectrum Management for QRD-Based Precoding with User Encoding Ordering in Downstream Transmission

In next generation DSL networks such as, employing discrete multi-tone transmission in high frequencies up to 212 MHz, the crosstalk among lines reaches very high levels. To precompensate the crosstalk in downstream transmission, QRD-based precoding has been proposed as a near-optimal dynamic spectrum management (DSM) technique. However, the performance ...

January 01, 2018

Practical Mitigation of Passive Intermodulation in Microstrip Circuits

This paper presents new experimental evidence and a novel practical approach for mitigation of passive intermodulation (PIM) in microstrip circuits fabricated on commercial printed circuit board laminates. The mechanisms of distributed PIM in microstrip circuits are reviewed and a phenomenology of PIM generation due to locally enhanced electromagnetic fields at ...

January 01, 2018

Efficient Cooperative HARQ for Multi-Source Multi-Relay Wireless Networks

In this paper, we compare the performance of three different cooperative Hybrid Automatic Repeat reQuest (HARQ) protocols for slow-fading half-duplex orthogonal multiple access multiple relay channel. Channel State Information (CSI) is available at the receiving side of each link only. Time Division Multiplexing is assumed, where each orthogonal transmission occurs ...