MobiSR

Lee, Royson; Venieris, Stylianos I.; Dudziak, Łukasz; Bhattacharya, Sourav; Lane, Nicholas D.

doi:10.1145/3300061.3345455

Cited by 58 publications

(3 citation statements)

References 51 publications

(72 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These techniques include skipping over parts of the inference graph, dynamically pruning the graph, or fractionally executing certain layers. Model selection or cascades [61], [47], [28], [24] are a type of adaptive inference where a family of models (with varying computational requirements) is trained for the same task. At inference time, either the appropriate model for the given input is selected or the input is propagated through more complex models until a certain criteria is met.…”

Section: Preliminariesmentioning

confidence: 99%

Timing Channels in Adaptive Neural Networks

Akinsanya,

Brennan

2024

Proceedings 2024 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

Current machine learning systems offer great predictive power but also require significant computational resources. As a result, the promise of a class of optimized machine learning models, called adaptive neural networks (ADNNs), has seen recent wide appeal. These models make dynamic decisions about the amount of computation to perform based on the given input, allowing for fast predictions on "easy" input. While various considerations of ADNNs have been extensively researched, how these input-dependent optimizations might introduce vulnerabilities has been hitherto under-explored. Our work is the first to demonstrate and evaluate timing channels due to the optimizations of ADNNs with the capacity to leak sensitive attributes about a user's input. We empirically study six ADNNs types and demonstrate how an attacker can significantly improve their ability to infer sensitive attributes, such as class label, of another user's input from an observed timing measurement. Our results show that timing information can increase an attacker's probability of correctly inferring the attribute of the user's input by up to a factor of 9.89x. Our empirical evaluation uses four different datasets, including those containing sensitive medical and demographic information, and considers leakage across a variety of sensitive attributes of the user's input. We conclude by demonstrating how timing channels can be exploited across the public internet in two fictitious web applications -Fictitious Health Company and Fictitious HR -that make use of ADNNs for serving predictions to their clients.

show abstract

Section: Preliminariesmentioning

confidence: 99%

Timing Channels in Adaptive Neural Networks

Akinsanya,

Brennan

2024

Proceedings 2024 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

show abstract

“…Deep Neural Networks are enabling several of the most exciting and innovative applications that are executed on a variety of computing devices, ranging from servers to edge and mobile devices. From a systems research viewpoint, this had led to a large set of ongoing projects on optimizing DNN inference (and training) tasks [1,21,23,26,32,34,61,66,69] as well as tensor compilers [31,33,54].…”

Section: Introductionmentioning

confidence: 99%

SoD ² : Statically Optimizing Dynamic Deep Neural Network Execution

Niu,

Agrawal,

Ren

2024

Proceedings of the 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems,

View full text Add to dashboard Cite

Though many compilation and runtime systems have been developed for DNNs in recent years, the focus has largely been on static DNNs. Dynamic DNNs, where tensor shapes and sizes and even the set of operators used are dependent upon the input and/or execution, are becoming common. This paper presents SoD 2 , a comprehensive framework for optimizing Dynamic DNNs. The basis of our approach is a classification of common operators that form DNNs, and the use of this classification towards a Rank and Dimension Propagation (RDP) method. This framework statically determines the shapes of operators as known constants, symbolic constants, or operations on these. Next, using RDP we enable a series of optimizations, like fused code generation, execution (order) planning, and even runtime memory allocation plan generation. By evaluating the framework on 10 emerging Dynamic DNNs and comparing it against several existing systems, we demonstrate both reductions in execution latency and memory requirements, with RDP-enabled key optimizations responsible for much of the gains. Our evaluation results show that SoD 2 runs up to 3.9× faster than these systems while saving up to 88% peak memory consumption. CCS CONCEPTS• Computing methodologies → Neural networks; • Software and its engineering → Source code generation; • Human-centered computing → Mobile computing.

show abstract

“…Among this huge family, an emerging representative [13,16,31,38] studies the prospect of utilizing SR model to upscale the resolution of the LR video in lieu of transmitting the HR video directly, which in many cases, consumes tremendous bandwidth between servers and clients [19]. One practical method is to deploy a pretrained SR model on the devices of the end users [25,54], and perform resolution upscaling for the transmitted LR videos, thus obtaining HR videos without causing bandwidth congestion. However, the deployed SR model that is trained with limited data usually suffers from limited generalization abil- ity, and may not achieve good performance at the presence of new data distribution [55].…”

Section: Introductionmentioning

confidence: 99%

Towards High-Quality and Efficient Video Super-Resolution via Spatial-Temporal Data Overfitting

Li¹,

Ji²,

Qin³

et al. 2023

Preprint

View full text Add to dashboard Cite

As deep convolutional neural networks (DNNs) are widely used in various fields of computer vision, leveraging the overfitting ability of the DNN to achieve video resolution upscaling has become a new trend in the modern video delivery system. By dividing videos into chunks and overfitting each chunk with a super-resolution model, the server encodes videos before transmitting them to the clients, thus achieving better video quality and transmission efficiency. However, a large number of chunks are expected to ensure good overfitting quality, which substantially increases the storage and consumes more bandwidth resources for data transmission. On the other hand, decreasing the number of chunks through training optimization techniques usually requires high model capacity, which significantly slows down execution speed. To reconcile such, we propose a novel method for high-quality and efficient video resolution upscaling tasks, which leverages the spatial-temporal information to accurately divide video into chunks, thus keeping the number of chunks as well as the model size to minimum. Additionally, we advance our method into a single overfitting model by a data-aware joint training technique, which further reduces the storage requirement with negligible quality drop. We deploy our models on an offthe-shelf mobile phone, and experimental results show that our method achieves real-time video super-resolution with high video quality. Compared with the state-of-the-art, our method achieves 28 fps streaming speed with 41.6 PSNR, which is 14× faster and 2.29 dB better in the live video resolution upscaling tasks. Our codes are available in here.

show abstract

MobiSR

Cited by 58 publications

References 51 publications

Timing Channels in Adaptive Neural Networks

Timing Channels in Adaptive Neural Networks

SoD ² : Statically Optimizing Dynamic Deep Neural Network Execution

Towards High-Quality and Efficient Video Super-Resolution via Spatial-Temporal Data Overfitting

Contact Info

Product

Resources

About

MobiSR

Cited by 58 publications

References 51 publications

Timing Channels in Adaptive Neural Networks

Timing Channels in Adaptive Neural Networks

SoD 2 : Statically Optimizing Dynamic Deep Neural Network Execution

Towards High-Quality and Efficient Video Super-Resolution via Spatial-Temporal Data Overfitting

Contact Info

Product

Resources

About

SoD ² : Statically Optimizing Dynamic Deep Neural Network Execution