Filtering by
- All Subjects: artificial intelligence
- Creators: Panchanathan, Sethuraman
- Creators: Dean, W.P. Carey School of Business
- Resource Type: Text
Generative models have recently gained popularity as they synthesize unseen visual features and convert zero-shot learning into a classical supervised learning problem. These generative models are trained using seen classes and are expected to implicitly transfer the knowledge from seen to unseen classes. However, their performance is stymied by overfitting towards seen classes, which leads to substandard performance in generalized zero-shot learning. To address this concern, this dissertation proposes a novel generative model that leverages the semantic relationship between seen and unseen categories and explicitly performs knowledge transfer from seen categories to unseen categories. Experiments were conducted on several benchmark datasets to demonstrate the efficacy of the proposed model for both zero-shot learning and generalized zero-shot learning. The dissertation also provides a unique Student-Teacher based generative model for zero-shot learning and concludes with future research directions in this area.
The sudden turn to artificial intelligence has been widely supported because of the several proposed positive outcomes of using such technologies to support or replace humans. Automating tedious processes and removing potential human error is exciting for society, but some concerns must be addressed. This essay aims to understand how artificial intelligence can automate domains that likely significantly impact underprivileged and underrepresented groups. This essay will address the potentially devastating effects of algorithmic biases and AI’s contribution to perpetual economic inequality by surveying different domains, such as the justice system and the real estate industry. Without society broadly understanding the potential negative side effects on systems that matter, the rapid growth of artificial intelligence is a recipe for disaster. Everyone must become educated about AI’s current and potential implications before it is too late to stop its damaging effects.