• Publications
  • Influence
Adaptive deep learning model selection on embedded systems
TLDR
This paper presents an adaptive scheme to determine which DNN model to use for a given input, by considering the desired accuracy and inference time. Expand
  • 48
  • 2
  • PDF
Optimizing Deep Learning Inference on Embedded Systems Through Adaptive Model Selection
TLDR
We propose a new, alternative approach to enable efficient execution of DNNs on embedded devices. Expand
  • 14
  • 2
  • PDF
Adaptive optimization for OpenCL programs on embedded heterogeneous systems
TLDR
We present an automatic approach to map OpenCL kernels onto heterogeneous multi-cores for a given optimization criterion – whether it is faster runtime, lower energy consumption or a trade-off between them. Expand
  • 41
  • 1
  • PDF
Improving spark application throughput via memory aware task co-location: a mixture of experts approach
TLDR
This paper, we present a mixture-of-experts approach to model the memory behavior of Spark applications. Expand
  • 38
  • PDF
Adaptive Selection of Deep Learning Models on Embedded Systems
TLDR
The recent ground-breaking advances in deep learning networks ( DNNs ) make them attractive for embedded systems. Expand
  • 4
  • PDF
Real-Time Power Cycling in Video on Demand Data Centres Using Online Bayesian Prediction
TLDR
We present a novel real-time power-cycling architecture, supported by a media distribution approach and online prediction model, to automatically determine when servers are needed based on demand. Expand
  • 2
  • PDF
Efficient Leader Election for Synchronous Shared-Memory Systems
Leader election is a frequent problem for systems where it is important to coordinate activities of a group of actors. It has been extensively studied in the context of networked systems. But withExpand
  • 2
App Feature Extraction Model Calibration 1 2 3 4 Offline Profiling Runs Memory footprint Training programs Model Fitting Feature Extraction f Memory function Feature values Task Scheduling Func
Data analytic applications built upon big data processing frameworks such as Apache Spark are an important class of applications. Many of these applications are not latency-sensitive and thus can runExpand
AICO, Artificial Intelligent COach
TLDR
We have developed an artificial intelligent American football coach to help the coaches to determine the best defensive strategies to be used against an opponent. Expand
Adaptive optimization for OpenCL programs on embedded heterogeneous systems
Heterogeneous multi-core architectures consisting of CPUs and GPUs are commonplace in today’s embedded systems. These architectures offer potential for energy efficient computing if the applicationExpand