About the Project
Deep learning has become the dominant approach in the machine learning research community due to several astonishing results in the last two and a half years. Deep learning algorithms have been proposed to move machine learning systems towards the discovery of an abstract representation spread across multiple, hierarchical levels. Since 2012, there have been important empirical successes in the field of visual big data analytics starting with the success in the ImageNet challenge 2012 and a large boost in object detection performance in 2013 and 2014. Deep learning is attracting tremendous attention from the academic, industrial and media communities. Companies such as Google, Microsoft, Apple, IBM and Baidu are investing in deep learning and establishing research institutes in this exciting area.
Deep learning, especially deep neural networks, have shown outstanding performance on visual classification tasks and more recently on object localization. However, applying deep neural networks to general visual big data analytics and interpretation is still in its infantry and many open questions remain unanswered. This research project aims to investigate recent advances in state-of-the-art computer vision and machine learning theories, and study deep learning architectures to create a breakthrough in the field of visual big data processing and analysis, including deep learning in the context of video understanding, especially action recognition, activity analysis and pose estimation. The project will also include the development of sophisticated data augmentation techniques to generate big data for supervised deep learning and real-time deep learning applications in low-level vision tasks. Novel visual big data analytics techniques with deep learning, such as hierarchical visual feature learning and representation, new deep architectures that are suitable for image sequences, and deep feature representation transfer between multiple visual tasks, will be thoroughly studied in this project.
For further details of how to apply, entry requirements and the application form, see
https://www.northumbria.ac.uk/research/postgraduate-research-degrees/how-to-apply/
Please ensure you quote the advert reference above on your application form.
Deadline for applications: 31 October 2015
Start Date: 1 February 2016
References
F. Zhu and L. Shao, “Weakly-Supervised Cross-Domain Dictionary Learning for Visual Recognition”, International Journal of Computer Vision (IJCV), vol. 109, no. 1-2, pp. 42-59, Aug. 2014.
L. Shao, D. Wu and X. Li, “Learning Deep and Wide: A Spectral Method for Learning Deep Networks”, IEEE Transactions on Neural Networks and Learning Systems, vol. 25, no. 12, pp. 2303-2308, Dec. 2014.
L. Shao, L. Liu and X. Li, “Feature Learning for Image Classification via Multiobjective Genetic Programming”, IEEE Transactions on Neural Networks and Learning Systems, vol. 25, no. 7, pp. 1359-1371, Jul. 2014.
L. Liu, M. Yu and L. Shao, “Multiview Alignment Hashing for Efficient Image Search”, IEEE Transactions on Image Processing (2015), doi: 10.1109/TIP.2015.2390975.
L. Zhang, X. Zhen and L. Shao, “Learning Object-to-Class Kernels for Scene Classification”, IEEE Transactions on Image Processing, vol. 23, no. 8, pp. 3241-3253, Aug. 2014.
S. Jones and L. Shao, “Unsupervised Spectral Dual Assignment Clustering of Human Actions in Context”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, USA, 2014.
D. Wu and L. Shao, “Leveraging Hierarchical Parametric Networks for Skeletal Joints Based Action Segmentation and Recognition”, IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Columbus, USA, 2014.
F. Zhu, L. Shao and M. Yu, “Cross-Modality Submodular Dictionary Learning for Information Retrieval”, ACM International Conference on Information and Knowledge Management (CIKM), Shanghai, China, 2014.
L. Liu and L. Shao, “Learning Discriminative Representations from RGB-D Video Data”, International Joint Conference on Artificial Intelligence (IJCAI), Beijing, China, 2013.
Continue with Facebook
