The past decade has witnessed the rising dominance
of deep learning and artificial intelligence in a wide range of
applications. In particular, the ocean of wireless smartphones and
IoT devices continue to fuel the tremendous growth of edge/cloudbased
machine learning (ML) systems including image/speech
recognition and classification. To overcome the infrastructural
barrier of limited network bandwidth in cloud ML, existing
solutions have mainly relied on traditional compression codecs
such as JPEG that were historically engineered for humanend
users instead of ML algorithms. Traditional codecs do not
necessarily preserve features important to ML algorithms under
limited bandwidth, leading to potentially inferior performance.
This work investigates application-driven optimization of programmable
commercial codec settings for networked learning
tasks such as image classification. Based on the foundation
of variational autoencoders (VAEs), we develop an end-to-end
networked learning framework by jointly optimizing the codec
and classifier without reconstructing images for given data rate
(bandwidth). Compared with standard JPEG codec, the proposed
VAE joint compression and classification framework achieves
classification accuracy improvement by over 10% and 4%,
respectively, for CIFAR-10 and ImageNet-1k data sets at data rate
of 0.8 bpp. Our proposed VAE-based models show 65%99% reductions
in encoder size, 1.5 13.1 improvements in inference
speed and 25%99% savings in power compared to baseline
models. We further show that a simple decoder can reconstruct
images with sufficient quality without compromising classification
accuracy.
more »
« less
Event-Driven Deep Learning for Edge Intelligence (EDL-EI) †
Edge intelligence (EI) has received a lot of interest because it can reduce latency, increase efficiency, and preserve privacy. More significantly, as the Internet of Things (IoT) has proliferated, billions of portable and embedded devices have been interconnected, producing zillions of gigabytes on edge networks. Thus, there is an immediate need to push AI (artificial intelligence) breakthroughs within edge networks to achieve the full promise of edge data analytics. EI solutions have supported digital technology workloads and applications from the infrastructure level to edge networks; however, there are still many challenges with the heterogeneity of computational capabilities and the spread of information sources. We propose a novel event-driven deep-learning framework, called EDL-EI (event-driven deep learning for edge intelligence), via the design of a novel event model by defining events using correlation analysis with multiple sensors in real-world settings and incorporating multi-sensor fusion techniques, a transformation method for sensor streams into images, and lightweight 2-dimensional convolutional neural network (CNN) models. To demonstrate the feasibility of the EDL-EI framework, we presented an IoT-based prototype system that we developed with multiple sensors and edge devices. To verify the proposed framework, we have a case study of air-quality scenarios based on the benchmark data provided by the USA Environmental Protection Agency for the most polluted cities in South Korea and China. We have obtained outstanding predictive accuracy (97.65% and 97.19%) from two deep-learning models on the cities’ air-quality patterns. Furthermore, the air-quality changes from 2019 to 2020 have been analyzed to check the effects of the COVID-19 pandemic lockdown.
more »
« less
- NSF-PAR ID:
- 10294395
- Date Published:
- Journal Name:
- Sensors
- Volume:
- 21
- Issue:
- 18
- ISSN:
- 1424-8220
- Page Range / eLocation ID:
- 6023
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Evolution has honed predatory skills in the natural world where localizing and intercepting fast-moving prey is required. The current generation of robotic systems mimics these biological systems using deep learning. High-speed processing of the camera frames using convolutional neural networks (CNN) (frame pipeline) on such constrained aerial edge-robots gets resource-limited. Adding more compute resources also eventually limits the throughput at the frame rate of the camera as frame-only traditional systems fail to capture the detailed temporal dynamics of the environment. Bio-inspired event cameras and spiking neural networks (SNN) provide an asynchronous sensor-processor pair (event pipeline) capturing the continuous temporal details of the scene for high-speed but lag in terms of accuracy. In this work, we propose a target localization system combining event-camera and SNN-based high-speed target estimation and frame-based camera and CNN-driven reliable object detection by fusing complementary spatio-temporal prowess of event and frame pipelines. One of our main contributions involves the design of an SNN filter that borrows from the neural mechanism for ego-motion cancelation in houseflies. It fuses the vestibular sensors with the vision to cancel the activity corresponding to the predator's self-motion. We also integrate the neuro-inspired multi-pipeline processing with task-optimized multi-neuronal pathway structure in primates and insects. The system is validated to outperform CNN-only processing using prey-predator drone simulations in realistic 3D virtual environments. The system is then demonstrated in a real-world multi-drone set-up with emulated event data. Subsequently, we use recorded actual sensory data from multi-camera and inertial measurement unit (IMU) assembly to show desired working while tolerating the realistic noise in vision and IMU sensors. We analyze the design space to identify optimal parameters for spiking neurons, CNN models, and for checking their effect on the performance metrics of the fused system. Finally, we map the throughput controlling SNN and fusion network on edge-compatible Zynq-7000 FPGA to show a potential 264 outputs per second even at constrained resource availability. This work may open new research directions by coupling multiple sensing and processing modalities inspired by discoveries in neuroscience to break fundamental trade-offs in frame-based computer vision 1 .more » « less
-
null (Ed.)Human activity recognition (HAR) from wearable sensors data has become ubiquitous due to the widespread proliferation of IoT and wearable devices. However, recognizing human activity in heterogeneous environments, for example, with sensors of different models and make, across different persons and their on-body sensor placements introduces wide range discrepancies in the data distributions, and therefore, leads to an increased error margin. Transductive transfer learning techniques such as domain adaptation have been quite successful in mitigating the domain discrepancies between the source and target domain distributions without the costly target domain data annotations. However, little exploration has been done when multiple distinct source domains are present, and the optimum mapping to the target domain from each source is not apparent. In this paper, we propose a deep Multi-Source Adversarial Domain Adaptation (MSADA) framework that opportunistically helps select the most relevant feature representations from multiple source domains and establish such mappings to the target domain by learning the perplexity scores. We showcase that the learned mappings can actually reflect our prior knowledge on the semantic relationships between the domains, indicating that MSADA can be employed as a powerful tool for exploratory activity data analysis. We empirically demonstrate that our proposed multi-source domain adaptation approach achieves 2% improvement with OPPORTUNITY dataset (cross-person heterogeneity, 4 ADLs), whereas 13% improvement on DSADS dataset (cross-position heterogeneity, 10 ADLs and sports activities).more » « less
-
Mobile edge computing pushes computationally-intensive services closer to the user to provide reduced delay due to physical proximity. This has led many to consider deploying deep learning models on the edge – commonly known as edge intelligence (EI). EI services can have many model implementations that provide different QoS. For instance, one model can perform inference faster than another (thus reducing latency) while achieving less accuracy when evaluated. In this paper, we study joint service placement and model scheduling of EI services with the goal to maximize Quality-of-Servcice (QoS) for end users where EI services have multiple implementations to serve user requests, each with varying costs and QoS benefits. We cast the problem as an integer linear program and prove that it is NP-hard. We then prove the objective is equivalent to maximizing a monotone increasing, submodular set function and thus can be solved greedily while maintaining a (1 – 1/e)-approximation guarantee. We then propose two greedy algorithms: one that theoretically guarantees this approximation and another that empirically matches its performance with greater efficiency. Finally, we thoroughly evaluate the proposed algorithm for making placement and scheduling decisions in both synthetic and real-world scenarios against the optimal solution and some baselines. In the real-world case, we consider real machine learning models using the ImageNet 2012 data-set for requests. Our numerical experiments empirically show that our more efficient greedy algorithm is able to approximate the optimal solution with a 0.904 approximation on average, while the next closest baseline achieves a 0.607 approximation on average.more » « less
-
The management of drinking water quality is critical to public health and can benefit from techniques and technologies that support near real-time forecasting of lake and reservoir conditions. The cyberinfrastructure (CI) needed to support forecasting has to overcome multiple challenges, which include: 1) deploying sensors at the reservoir requires the CI to extend to the network’s edge and accommodate devices with constrained network and power; 2) different lakes need different sensor modalities, deployments, and calibrations; hence, the CI needs to be flexible and customizable to accommodate various deployments; and 3) the CI requires to be accessible and usable to various stakeholders (water managers, reservoir operators, and researchers) without barriers to entry. This paper describes the CI underlying FLARE (Forecasting Lake And Reservoir Ecosystems), a novel system co-designed in an interdisciplinary manner between CI and domain scientists to address the above challenges. FLARE integrates R packages that implement the core numerical forecasting (including lake process modeling and data assimilation) with containers, overlay virtual networks, object storage, versioned storage, and event-driven Function-as-a-Service (FaaS) serverless execution. It is a flexible forecasting system that can be deployed in different modalities, including the Manual Mode suitable for end-users’ personal computers and the Workflow Mode ideal for cloud deployment. The paper reports on experimental data and lessons learned from the operational deployment of FLARE in a drinking water supply (Falling Creek Reservoir in Vinton, Virginia, USA). Experiments with a FLARE deployment quantify its edge-to-cloud virtual network performance and serverless execution in OpenWhisk deployments on both XSEDE-Jetstream and the IBM Cloud Functions FaaS system.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3390/s21186023
