Haocheng Fang | Searcharxiv

arXiv2022

A generalized model of magnon kinetics and subgap magnetic noise

Magnetic noise spectroscopy provides a noninvasive probe of spin dynamics in magnetic materials. We consider two-dimensional magnetically ordered insulators with magnon excitations, especially those supporting long-distance magnon transport, where nitrogen-vacancy (NV) centers enable the access to (nearly) ballistic transport regime of magnons. We develop a generalized theory to describe the magnon transport across a wide range of length scales. The longitudinal dynamic spin susceptibility is derived from the Boltzmann equation and extended to a Lindhard form, which is modified by both the spin-conserving magnon collisions and spin relaxation. Our result is consistent with the diffusive (ballistic) model for the length scale much larger (smaller) than the magnon mean free path, and provides a description for the intermediate regime. We also give a prediction for the NV transition rate in different magnon transport regimes.

cond-mat.mes-hall↗

arXiv2018

Designing Adaptive Neural Networks for Energy-Constrained Image Classification

As convolutional neural networks (CNNs) enable state-of-the-art computer vision applications, their high energy consumption has emerged as a key impediment to their deployment on embedded and mobile devices. Towards efficient image classification under hardware constraints, prior work has proposed adaptive CNNs, i.e., systems of networks with different accuracy and computation characteristics, where a selection scheme adaptively selects the network to be evaluated for each input image. While previous efforts have investigated different network selection schemes, we find that they do not necessarily result in energy savings when deployed on mobile systems. The key limitation of existing methods is that they learn only how data should be processed among the CNNs and not the network architectures, with each network being treated as a blackbox. To address this limitation, we pursue a more powerful design paradigm where the architecture settings of the CNNs are treated as hyper-parameters to be globally optimized. We cast the design of adaptive CNNs as a hyper-parameter optimization problem with respect to energy, accuracy, and communication constraints imposed by the mobile device. To efficiently solve this problem, we adapt Bayesian optimization to the properties of the design space, reaching near-optimal configurations in few tens of function evaluations. Our method reduces the energy consumed for image classification on a mobile device by up to 6x, compared to the best previously published work that uses CNNs as blackboxes. Finally, we evaluate two image classification practices, i.e., classifying all images locally versus over the cloud under energy and communication constraints.

cs.LG↗