Document Type
Doctoral Thesis
Disciplines
1.2 COMPUTER AND INFORMATION SCIENCE
Abstract
Deep learning has developed rapidly since the introduction of Deep Belief Networks during the past decade. As an area of machine learning, it still has many open challenges. Among these open challenges is the issue of transparency, with deep learning models known as black boxes. Both explainability of a model for understanding the decision making process, and the transparency of the relationship between the input and output are crucial for understanding a model. The understanding of models can build trust between AI systems and humans, verify models behavior, and identify potential biases or errors.
DOI
https://doi.org/10.21427/cq9a-xj52
Recommended Citation
Liu, Xuehao, "CNNs and Transformers for Visual Understanding: From Feature Alignment to Image Captioning" (2025). Doctoral. 3.
https://arrow.tudublin.ie/compdidadoc/3
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License.
Publication Details
Thesis submitted for the degree of Doctor of Philosophy, Technological University Dublin, 2025.
doi:10.21427/cq9a-xj52