Open Access. Powered by Scholars. Published by Universities.®
- Keyword
-
- Deep learning (2)
- Medical imaging (2)
- Visual perception (2)
- Action detection (1)
- Activation maximization (1)
-
- Automated driving (1)
- Automated parking (1)
- Colourisation (1)
- Computational photography (1)
- Conditional-generative adversarial networks (1)
- Deep learning interpretability (1)
- Embedded vision (1)
- FAST super-resolution processing (1)
- Fisheye cameras (1)
- GAN (1)
- Game RoShambo (1)
- GradCam (1)
- Image processing applications (1)
- Moving object detection (1)
- Neuromorphic processing (1)
- Object detection (1)
- PET-CT bone marrow (1)
- Segmentation (1)
- Shape analysis (1)
- Space-time (1)
- Time-lapse video generation (1)
- Timestamp removal (1)
- Volume approximation (1)
Articles 1 - 8 of 8
Full-Text Articles in Engineering
Visualizing And Interpreting Feature Reuse Of Pretrained Cnns For Histopathology, Mara Graziani, Vincent Andrearczyk, Henning Muller
Visualizing And Interpreting Feature Reuse Of Pretrained Cnns For Histopathology, Mara Graziani, Vincent Andrearczyk, Henning Muller
Session 6: Applications, Architecture and Systems Integration
Reusing the parameters of networks pretrained on large scale datasets of natural images, such as ImageNet, is a common technique in the medical imaging domain. The large variability of objects and classes is, however, drastically reduced in most medical applications where images are dominated by repetitive patterns with, at times, subtle differences between the classes. This paper takes the example of finetuning a pretrained convolutional network on a histopathology task. Because of the reduced visual variability in this application domain, the network mostly learns to detect textures and simple patterns. As a result, the complex structures that maximize the channel …
Spatial Coherency In Colourisation, Sean Mullery, Paul F. Whelan
Spatial Coherency In Colourisation, Sean Mullery, Paul F. Whelan
Session 6: Applications, Architecture and Systems Integration
Automatic colourisation is the function of inferring colour information from a grey-scale prior and then combining the colour with the grey-scale to form a colourised version of the image. We identify Spatial Coherence as a particular weakness in methods that use Convolutional Neural Networks for colourisation. Generated colours do not adhere to semantic edges and are not consistent within boundaries where we would expect uniform colour. Spatial Coherence, while often evident to the human eye, does not yet have an objective metric. We show, by segmentation of the combined ab channels of the CIEL*a*b* colour space, that a segmentation based …
Fisheyemodnet: Moving Object Detection On Surround-View Cameras For Autonomous Driving, Marie Yahiaoui, Hazem Rashed, Letizia Mariotti, Ganesh Sistu, Ian Clancy, Lucie Yahiaoui, Senthil Yogamani
Fisheyemodnet: Moving Object Detection On Surround-View Cameras For Autonomous Driving, Marie Yahiaoui, Hazem Rashed, Letizia Mariotti, Ganesh Sistu, Ian Clancy, Lucie Yahiaoui, Senthil Yogamani
Session 6: Applications, Architecture and Systems Integration
Moving Object Detection (MOD) is an important task for achieving robust autonomous driving. An autonomous vehicle has to estimate collision risk with other interacting objects in the environment and calculate an optional trajectory. Collision risk is typically higher for moving objects than static ones due to the need to estimate the future states and poses of the objects for decision making. This is particularly important for near-range objects around the vehicle which are typically detected by a fisheye surroundview system that captures a 360± view of the scene. In this work, we propose a CNN architecture for moving object detection …
Neuromorphic Event-Based Action Recognition, S. Harrigan, S. Colman, D. Kerr, P. Yogarajah, Z. Fang, C. Wu
Neuromorphic Event-Based Action Recognition, S. Harrigan, S. Colman, D. Kerr, P. Yogarajah, Z. Fang, C. Wu
Session 6: Applications, Architecture and Systems Integration
An action can be viewed as spike trains or streams of events when observed and captured by neuromorphic imaging hardware such as the iniLabs DVS128. These streams are unique to each action enabling them to be used to form descriptors. This paper describes an approach for detecting specific actions based on space-time template matching by forming such descriptors and using them as comparative tools. The developed approach is used to detect symbols from the popular RoShambo (rock, paper and scissors) game. The results demonstrate that the developed approach can be used to correctly detect the motions involved in producing RoShambo …
Synthetic Positron Emission Tomography Using Conditional-Generative Adversarial Networks For Healthy Bone Marrow Baseline Image Generation, Patrick Leydon, Martin O'Connell, Derek Greene, Kathleen Curran
Synthetic Positron Emission Tomography Using Conditional-Generative Adversarial Networks For Healthy Bone Marrow Baseline Image Generation, Patrick Leydon, Martin O'Connell, Derek Greene, Kathleen Curran
Session 6: Applications, Architecture and Systems Integration
A Conditional-Generative Adversarial Network has been used for a supervised image-to-image transla- tion task which outputs a synthetic PET scan based on real patient CT data. The network is trained using only data of patients with healthy bone marrow metabolism. This allows for a patient specific synthetic healthy baseline scan to be produced. This can be used by a clinician for comparison to real PET data in the absence of a baseline scan or to aid in the diagnosis of conditions such as Multiple Myeloma which manifest as changes in bone marrow metabolism.
Development Of A Nanodrop Shape Analysis Tool For Installation In A Novel Nanodrop Spectrophotometer, Colin Monaghan, Jane Courtney
Development Of A Nanodrop Shape Analysis Tool For Installation In A Novel Nanodrop Spectrophotometer, Colin Monaghan, Jane Courtney
Session 6: Applications, Architecture and Systems Integration
The rapid identification of liquid composition is an important task integral to a wide range of industries including medical, pharmaceuticals, petrochemicals, and vinification. To aid in this identification spectroscopy can be utilised, however specialised instrumentation must be developed to deliver quantitative information. A spectrophotometer uses spectral data to identify chemical composition of droplets. However, to accurately perform this function, prior knowledge of the size and shape of the droplet is essential to understand chemical quantity. Whilst image data can be easily captured with a high definition camera, the image analysis to translate images into a relevant region of interest (ROI) …
Fisheyemultinet: Real-Time Multi-Task Learning Architecture For Surround-View Automated Parking System., Pullaro Maddu, Wayne Doherty, Ganesh Sistu, Isabelle Leang, Michal Uricar, Sumanth Chennupati, Hazem Rashed, Jonathan Horgan, Ciaran Hughes, Senthil Yogamani
Fisheyemultinet: Real-Time Multi-Task Learning Architecture For Surround-View Automated Parking System., Pullaro Maddu, Wayne Doherty, Ganesh Sistu, Isabelle Leang, Michal Uricar, Sumanth Chennupati, Hazem Rashed, Jonathan Horgan, Ciaran Hughes, Senthil Yogamani
Session 6: Applications, Architecture and Systems Integration
Automated Parking is a low speed manoeuvring scenario which is quite unstructured and complex, requiring full 360° near-field sensing around the vehicle. In this paper, we discuss the design and implementation of an automated parking system from the perspective of camera based deep learning algorithms. We provide a holistic overview of an industrial system covering the embedded system, use cases and the deep learning architecture. We demonstrate a real-time multi-task deep learning network called FisheyeMultiNet, which detects all the necessary objects for parking on a low-power embedded system. FisheyeMultiNet runs at 15 fps for 4 cameras and it has three …
An Efficient Approach To Automatic Generation Of Time-Lapse Video Sequence, Javier Calero De Torres, Bryan Gardiner, Ilias Dahi, Sandra Moffett, Marco Herbst, Joan Condell
An Efficient Approach To Automatic Generation Of Time-Lapse Video Sequence, Javier Calero De Torres, Bryan Gardiner, Ilias Dahi, Sandra Moffett, Marco Herbst, Joan Condell
Session 6: Applications, Architecture and Systems Integration
Time-lapse video sequences have recently become a highly utilised asset for marketing and advertising, particularly within the field of construction and landscape development. However, the manual generation of these videos, at a quality that can be used for marketing purposes, can be quite time-consuming. In this paper, a novel application for generating time-lapse videos is proposed, which will automatically select the optimal frames for time-lapse video generation, enhance these frames by applying a number of image pre- processing and machine learning techniques such as FAST super-resolution to improve the frames quality, and finally, provide an intuitive user interface to allow …