DLVD 2014 - IEEE International Workshop on Deep Learning on Visual Data

Deep learning is a set of algorithms in machine learning that attempts to model high-level abstractions in data by using architectures composed of multiple non-linear transformations. As an emerging AI technique, deep learning has achieved very impressive results and significantly outperformed state-of-the-art on various research topics. Deep learning had a strong impact on speech recognition and natural language processing, with breaking through results obtained by several academics as well as researchers at industrial labs, e.g. Microsoft, NEC, Google, bringing them to into products. Computer vision researchers have successfully used deep learning for a large amount of visual data processing and understanding problems such as object recognition, object detection, face recognition, action recognition, segmentation, pose estimation, denoising, and image quality assessment. The success comes from the capability of automatically learning hierarchical feature representations from data, disentangling hidden factors, and jointly optimizing functional layers. The recent research advances in deep learning, the explosive increase in the amount of available visual data and the computational capacities have provided us with unprecedented opportunities in processing and understanding visual data using deep learning.
Nevertheless, many problems have yet to be explored. For example, is there any new application of deep learning to visual data? How to formulate an image and video understanding problem with deep model? How to make use the past research experience on training and designing existing computer vision systems to design new deep models and training strategies in various applications? What are the new challenges when applying deep model to visual data?
This workshop provides a forum for researchers on deep learning and visual data processing to explore the new opportunities and challenges in related topics. This workshop targets on audiences who are interested in deep learning as well as visual data processing and understanding.