Explanipedia

Beyond The Rainbow: High Performance Deep Reinforcement Learning on a Desktop PC Open

T. Alan Clark, Mark Towers, Christine Evers, Jonathon Hare · 2024

Rainbow Deep Q-Network (DQN) demonstrated combining multiple independent enhancements could significantly boost a reinforcement learning (RL) agent's performance. In this paper, we present "Beyond The Rainbow" (BTR), a novel algorithm that…

Rethinking Deep Thinking: Stable Learning of Algorithms using Lipschitz Constraints Open

Javona White Bear, Adam Prügel‐Bennett, Jonathon Hare · 2024

Iterative algorithms solve problems by taking steps until a solution is reached. Models in the form of Deep Thinking (DT) networks have been demonstrated to learn iterative algorithms in a way that can scale to different sized problems at …

Fluid Dynamic DNNs for Reliable and Adaptive Distributed Inference on Edge Devices Open

Xun Lei, Mingyu Hu, Hengrui Zhao, Amit Kumar Singh, Jonathon Hare , et al. · 2024

Distributed inference is a popular approach for efficient DNN inference at the edge. However, traditional Static and Dynamic DNNs are not distribution-friendly, causing system reliability and adaptability issues. In this paper, we introduc…

Dynamic DNNs and Runtime Management for Efficient Inference on Mobile/Embedded Devices Open

Xun Lei, Jonathon Hare, Geoff V. Merrett · 2024

Deep neural network (DNN) inference is increasingly being executed on mobile and embedded platforms due to several key advantages in latency, privacy and always-on availability. However, due to limited computing resources, efficient DNN de…

Temporal Early Exits for Efficient Video Object Detection Open

Amin Sabet, Xun Lei, Jonathon Hare, Bashir M. Al‐Hashimi, Geoff V. Merrett · 2023

Efficiently transferring image-based object detectors to the domain of video remains challenging under resource constraints. Previous efforts used feature propagation to avoid recomputing unchanged features. However, the overhead is signif…

Temporal Early Exits for Efficient Video Object Detection Open

Amin Sabet, Xun Lei, Jonathon Hare, Bashir M. Al‐Hashimi, Geoff V. Merrett · 2023

Efficiently transferring image-based object detectors to the domain of video remains challenging under resource constraints. Previous efforts used feature propagation to avoid recomputing unchanged features. However, the overhead is signif…

Improving the Robustness of Neural Multiplication Units with Reversible Stochasticity Open

Bhumika Mistry, Katayoun Farrahi, Jonathon Hare · 2022

Multilayer Perceptrons struggle to learn certain simple arithmetic tasks. Specialist neural modules for arithmetic can outperform classical architectures with gains in extrapolation, interpretability and convergence speeds, but are highly …

Dynamic DNNs Meet Runtime Resource Management on Mobile and Embedded Platforms Open

Xun Lei, Bashir M. Al‐Hashimi, Jonathon Hare, Geoff V. Merrett · 2022

Deep neural network (DNN) inference is increasingly being executed on mobile and embedded platforms due to low latency and better privacy. However, efficient deployment on these platforms is challenging due to the intensive computation and…

Physically Embodied Deep Image Optimisation Open

Daniela Mihai, Jonathon Hare · 2022

Physical sketches are created by learning programs to control a drawing robot. A differentiable rasteriser is used to optimise sets of drawing strokes to match an input image, using deep networks to provide an encoding for which we can com…

Temporal Early Exits for Efficient Video Object Detection Open

Amin Sabet, Jonathon Hare, Bashir M. Al‐Hashimi, Geoff V. Merrett · 2022

Temporal Early Exits for Efficient Video Object Detection Open

Amin Sabet, Jonathon Hare, Bashir M. Al‐Hashimi, Geoff V. Merrett · 2022

Temporal Early Exits for Efficient Video Object Detection Open

Amin Sabet, Jonathon Hare, Bashir M. Al‐Hashimi, Geoff V. Merrett · 2022

Image-based attitude determination of co-orbiting satellites using deep learning technologies Open

Ben Guthrie, Minkwan Kim, Hodei Urrutxua, Jonathon Hare · 2021

Shared Visual Representations of Drawing for Communication: How do\n different biases affect human interpretability and intent? Open

Daniela Mihai, Jonathon Hare · 2021

We present an investigation into how representational losses can affect the\ndrawings produced by artificial agents playing a communication game. Building\nupon recent advances, we show that a combination of powerful pretrained encoder\nne…

Shared Visual Representations of Drawing for Communication: How do different biases affect human interpretability and intent? Open

Daniela Mihai, Jonathon Hare · 2021

We present an investigation into how representational losses can affect the drawings produced by artificial agents playing a communication game. Building upon recent advances, we show that a combination of powerful pretrained encoder netwo…

Learning Division with Neural Arithmetic Logic Modules Open

Bhumika Mistry, Katayoun Farrahi, Jonathon Hare · 2021

To achieve systematic generalisation, it first makes sense to master simple tasks such as arithmetic. Of the four fundamental arithmetic operations (+,-,$\times$,$÷$), division is considered the most difficult for both humans and computers…

GhostShiftAddNet: More Features from Energy-Efficient Operations Open

Jia Bi, Jonathon Hare, Geoff V. Merrett · 2021

Deep convolutional neural networks (CNNs) are computationally and memory intensive. In CNNs, intensive multiplication can have resource implications that may challenge the ability for effective deployment of inference on resource-constrain…

Dynamic Transformer for Efficient Machine Translation on Embedded Devices Open

Hishan Parry, Xun Lei, Amin Sabet, Jia Bi, Jonathon Hare , et al. · 2021

The Transformer architecture is widely used for machine translation tasks. However, its resource-intensive nature makes it challenging to implement on constrained embedded devices, particularly where available hardware resources can vary a…

Dataset for "Dynamic Transformer for Efficient Machine Translation on Embedded Devices" Open

Hishan Parry, Xun Lei, Mohammadamin Sabetsarvestani, Jia Bi, Jonathon Hare , et al. · 2021

This dataset supports the publication: 'Dynamic Transformer for Efficient Machine Translation on Embedded Devices' in '3rd ACM/IEEE Workshop on Machine Learning for CAD (MLCAD'21)'.

Language Models as Zero-shot Visual Semantic Learners Open

Yue Jiao, Jonathon Hare, Adam Prügel‐Bennett · 2021

Visual Semantic Embedding (VSE) models, which map images into a rich semantic embedding space, have been a milestone in object recognition and zero-shot learning. Current approaches to VSE heavily rely on static word em-bedding techniques.…

What Remains of Visual Semantic Embeddings Open

Yue Jiao, Jonathon Hare, Adam Prügel‐Bennett · 2021

Zero shot learning (ZSL) has seen a surge in interest over the decade for its tight links with the mechanism making young children recognize novel objects. Although different paradigms of visual semantic embedding models are designed to al…

Dynamic Transformer for Efficient Machine Translation on Embedded\n Devices Open

Hishan Parry, Xun Lei, Amin Sabet, Jia Bi, Jonathon Hare , et al. · 2021

The Transformer architecture is widely used for machine translation tasks.\nHowever, its resource-intensive nature makes it challenging to implement on\nconstrained embedded devices, particularly where available hardware resources\ncan var…

Temporal Early Exits for Efficient Video Object Detection Open

Amin Sabet, Jonathon Hare, Bashir M. Al‐Hashimi, Geoff V. Merrett · 2021

Transferring image-based object detectors to the domain of video remains challenging under resource constraints. Previous efforts utilised optical flow to allow unchanged features to be propagated, however, the overhead is considerable whe…

Runtime DNN performance scaling through resource management on heterogeneous embedded platforms Open

Xun Lei, Bashir M. Al‐Hashimi, Jonathon Hare, Geoff V. Merrett · 2021

DNN inference is increasingly being executed locally on embedded platforms, due to the clear advantages in latency, privacy and connectivity. Modern SoCs typically execute a combination of different and dynamic workloads concurrently, it i…

Learning to Draw: Emergent Communication through Sketching Open

Daniela Mihai, Jonathon Hare · 2021

Evidence that visual communication preceded written language and provided a basis for it goes back to prehistory, in forms such as cave and rock paintings depicting traces of our distant ancestors. Emergent communication research has sough…

Dynamic-OFA: Runtime DNN Architecture Switching for Performance Scaling on Heterogeneous Embedded Platforms Open

Wei Lou, Xun Lei, Amin Sabet, Jia Bi, Jonathon Hare , et al. · 2021

Mobile and embedded platforms are increasingly required to efficiently execute computationally demanding DNNs across heterogeneous processing elements. At runtime, the available hardware resources to DNNs can vary considerably due to other…

Dynamic-OFA: Runtime DNN Architecture Switching for Performance Scaling on Heterogeneous Embedded Platforms Open

Wei Lou, Xun Lei, Amin Sabet, Jia Bi, Jonathon Hare , et al. · 2021

Mobile and embedded platforms are increasingly required to efficiently execute computationally demanding DNNs across heterogeneous processing elements. At runtime, the available hardware resources to DNNs can vary considerably due to other…

Dataset for "Dynamic-OFA: Runtime DNN Architecture Switching for Performance Scaling on Heterogeneous Embedded Platforms" Open

Wei Lou, Xun Lei, Mohammadamin Sabetsarvestani, Jia Bi, Jonathon Hare , et al. · 2021

This dataset supports the publication: 'Dynamic-OFA: Runtime DNN Architecture Switching for Performance Scaling on Heterogeneous Embedded Platforms' in 'Efficient Deep Learning for Computer Vision Workshop at CVPR Conference 2021'.

Differentiable Drawing and Sketching Open

Daniela Mihai, Jonathon Hare · 2021

We present a bottom-up differentiable relaxation of the process of drawing points, lines and curves into a pixel raster. Our approach arises from the observation that rasterising a pixel in an image given parameters of a primitive can be r…

Open-domain Topic Identification of Out-of-domain Utterances using Wikipedia Open

Alexandry Augustin, Alexandros Papangelis, Margarita Kotti, Pavlos Vougiouklis, Jonathon Hare , et al. · 2021

Users of spoken dialogue systems (SDS) expect high quality interactions across a wide range of diverse topics. However, the implementation of SDS capable of responding to every conceivable user utterance in an informative way is a challeng…

Jonathon Hare YOU? Author Swipe