A set of methods to address computer vision problems has been developed. Video un- derstanding is an activate area of research in recent years. If one can accurately identify salient objects in a video sequence, these components can be used in information retrieval and scene analysis. This research started with the development of a course-to-fine frame- work to extract salient objects in video sequences. Previous work on image and video frame background modeling involved methods that ranged from simple and efficient to accurate but computationally complex. It will be shown in this research that the novel approach to implement object extraction is efficient and effective that outperforms the existing state-of-the-art methods. However, the drawback to this method is the inability to deal with non-rigid motion.

With the rapid development of artificial neural networks, deep learning approaches are explored as a solution to computer vision problems in general. Focusing on image and text, the image (or video frame) understanding can be achieved using CVS. With this concept, modality generation and other relevant applications such as automatic im- age description, text paraphrasing, can be explored. Specifically, video sequences can be modeled by Recurrent Neural Networks (RNN), the greater depth of the RNN leads to smaller error, but that makes the gradient in the network unstable during training.To overcome this problem, a Batch-Normalized Recurrent Highway Network (BNRHN) was developed and tested on the image captioning (image-to-text) task. In BNRHN, the highway layers are incorporated with batch normalization which diminish the gradient vanishing and exploding problem. In addition, a sentence to vector encoding framework that is suitable for advanced natural language processing is developed. This semantic text embedding makes use of the encoder-decoder model which is trained on sentence paraphrase pairs (text-to-text). With this scheme, the latent representation of the text is shown to encode sentences with common semantic information with similar vector rep- resentations. In addition to image-to-text and text-to-text, an image generation model is developed to generate image from text (text-to-image) or another image (image-to- image) based on the semantics of the content. The developed model, which refers to the Multi-Modal Vector Representation (MMVR), builds and encodes different modalities into a common vector space that achieve the goal of keeping semantics and conversion between text and image bidirectional. The concept of CVS is introduced in this research to deal with multi-modal conversion problems. In theory, this method works not only on text and image, but also can be generalized to other modalities, such as video and audio. The characteristics and performance are supported by both theoretical analysis and experimental results. Interestingly, the MMVR model is one of the many possible ways to build CVS. In the final stages of this research, a simple and straightforward framework to build CVS, which is considered as an alternative to the MMVR model, is presented.

Publication Date


Document Type


Student Type


Degree Name

Imaging Science (Ph.D.)

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)


Carl Salvaggio

Advisor/Committee Member

Pengcheng Shi

Advisor/Committee Member

Raymond Ptucha


RIT – Main Campus