-
MorpheusNet: Resource efficient sleep stage classifier for embedded on-line systems
Authors:
Ali Kavoosi,
Morgan P. Mitchell,
Raveen Kariyawasam,
John E. Fleming,
Penny Lewis,
Heidi Johansen-Berg,
Hayriye Cagnan,
Timothy Denison
Abstract:
Sleep Stage Classification (SSC) is a labor-intensive task, requiring experts to examine hours of electrophysiological recordings for manual classification. This is a limiting factor when it comes to leveraging sleep stages for therapeutic purposes. With increasing affordability and expansion of wearable devices, automating SSC may enable deployment of sleep-based therapies at scale. Deep Learning…
▽ More
Sleep Stage Classification (SSC) is a labor-intensive task, requiring experts to examine hours of electrophysiological recordings for manual classification. This is a limiting factor when it comes to leveraging sleep stages for therapeutic purposes. With increasing affordability and expansion of wearable devices, automating SSC may enable deployment of sleep-based therapies at scale. Deep Learning has gained increasing attention as a potential method to automate this process. Previous research has shown accuracy comparable to manual expert scores. However, previous approaches require sizable amount of memory and computational resources. This constrains the ability to classify in real time and deploy models on the edge. To address this gap, we aim to provide a model capable of predicting sleep stages in real-time, without requiring access to external computational sources (e.g., mobile phone, cloud). The algorithm is power efficient to enable use on embedded battery powered systems. Our compact sleep stage classifier can be deployed on most off-the-shelf microcontrollers (MCU) with constrained hardware settings. This is due to the memory footprint of our approach requiring significantly fewer operations. The model was tested on three publicly available data bases and achieved performance comparable to the state of the art, whilst reducing model complexity by orders of magnitude (up to 280 times smaller compared to state of the art). We further optimized the model with quantization of parameters to 8 bits with only an average drop of 0.95% in accuracy. When implemented in firmware, the quantized model achieves a latency of 1.6 seconds on an Arm CortexM4 processor, allowing its use for on-line SSC-based therapies.
△ Less
Submitted 14 January, 2024;
originally announced January 2024.
-
A Deep Metric Learning Approach to Account Linking
Authors:
Aleem Khan,
Elizabeth Fleming,
Noah Schofield,
Marcus Bishop,
Nicholas Andrews
Abstract:
We consider the task of linking social media accounts that belong to the same author in an automated fashion on the basis of the content and metadata of their corresponding document streams. We focus on learning an embedding that maps variable-sized samples of user activity -- ranging from single posts to entire months of activity -- to a vector space, where samples by the same author map to nearb…
▽ More
We consider the task of linking social media accounts that belong to the same author in an automated fashion on the basis of the content and metadata of their corresponding document streams. We focus on learning an embedding that maps variable-sized samples of user activity -- ranging from single posts to entire months of activity -- to a vector space, where samples by the same author map to nearby points. The approach does not require human-annotated data for training purposes, which allows us to leverage large amounts of social media content. The proposed model outperforms several competitive baselines under a novel evaluation framework modeled after established recognition benchmarks in other domains. Our method achieves high linking accuracy, even with small samples from accounts not seen at training time, a prerequisite for practical applications of the proposed linking framework.
△ Less
Submitted 15 May, 2021;
originally announced May 2021.
-
Dangerous human-made interference with climate: A GISS modelE study
Authors:
J. Hansen,
M. Sato,
R. Ruedy,
P. Kharecha,
A. Lacis,
R. Miller,
L. Nazarenko,
K. Lo,
G. A. Schmidt,
G. Russell,
I. Aleinov,
S. Bauer,
E. Baum,
B. Cairns,
V. Canuto,
M. Chandler,
Y. Cheng,
A. Cohen,
A. Del Genio,
G. Faluvegi,
E. Fleming,
A. Friend,
T. Hall,
C. Jackman,
J. Jonas
, et al. (22 additional authors not shown)
Abstract:
We investigate the issue of "dangerous human-made interference with climate" using simulations with GISS modelE driven by measured or estimated forcings for 1880-2003 and extended to 2100 for IPCC greenhouse gas scenarios as well as the 'alternative' scenario of Hansen and Sato. Identification of 'dangerous' effects is partly subjective, but we find evidence that added global warming of more tha…
▽ More
We investigate the issue of "dangerous human-made interference with climate" using simulations with GISS modelE driven by measured or estimated forcings for 1880-2003 and extended to 2100 for IPCC greenhouse gas scenarios as well as the 'alternative' scenario of Hansen and Sato. Identification of 'dangerous' effects is partly subjective, but we find evidence that added global warming of more than 1 degree C above the level in 2000 has effects that may be highly disruptive. The alternative scenario, with peak added forcing ~1.5 W/m2 in 2100, keeps further global warming under 1 degree C if climate sensitivity is \~3 degrees C or less for doubled CO2. We discuss three specific sub-global topics: Arctic climate change, tropical storm intensification, and ice sheet stability. Growth of non-CO2 forcings has slowed in recent years, but CO2 emissions are now surging well above the alternative scenario. Prompt actions to slow CO2 emissions and decrease non-CO2 forcings are needed to achieve the low forcing of the alternative scenario.
△ Less
Submitted 16 October, 2006;
originally announced October 2006.
-
Climate simulations for 1880-2003 with GISS modelE
Authors:
J. Hansen,
M. Sato,
R. Ruedy,
P. Kharecha,
A. Lacis,
R. Miller,
L. Nazarenko,
K. Lo,
G. A. Schmidt,
G. Russell,
I. Aleinov,
S. Bauer,
E. Baum,
B. Cairns,
V. Canuto,
M. Chandler,
Y. Cheng,
A. Cohen,
A. Del Genio,
G. Faluvegi,
E. Fleming,
A. Friend,
T. Hall,
C. Jackman,
J. Jonas
, et al. (22 additional authors not shown)
Abstract:
We carry out climate simulations for 1880-2003 with GISS modelE driven by ten measured or estimated climate forcings. An ensemble of climate model runs is carried out for each forcing acting individually and for all forcing mechanisms acting together. We compare side-by-side simulated climate change for each forcing, all forcings, observations, unforced variability among model ensemble members,…
▽ More
We carry out climate simulations for 1880-2003 with GISS modelE driven by ten measured or estimated climate forcings. An ensemble of climate model runs is carried out for each forcing acting individually and for all forcing mechanisms acting together. We compare side-by-side simulated climate change for each forcing, all forcings, observations, unforced variability among model ensemble members, and, if available, observed variability. Discrepancies between observations and simulations with all forcings are due to model deficiencies, inaccurate or incomplete forcings, and imperfect observations. Although there are notable discrepancies between model and observations, the fidelity is sufficient to encourage use of the model for simulations of future climate change. By using a fixed well-documented model and accurately defining the 1880-2003 forcings, we aim to provide a benchmark against which the effect of improvements in the model, climate forcings, and observations can be tested. Principal model deficiencies include unrealistically weak tropical El Nino-like variability and a poor distribution of sea ice, with too much sea ice in the Northern Hemisphere and too little in the Southern Hemisphere. The greatest uncertainties in the forcings are the temporal and spatial variations of anthropogenic aerosols and their indirect effects on clouds.
△ Less
Submitted 11 April, 2007; v1 submitted 16 October, 2006;
originally announced October 2006.
