Object recognition is an important area in computer vision. Object recognition has been advanced significantly by deep learning that unifies feature extraction and classification. In general, deep neural networks, such as Convolution Neural Networks (CNNs), are trained in high-performance systems. Aiming to extend the reach of deep learning to personal computing, I propose a study of deep learning-based object recognition in low-end systems, such as laptops. This research includes how differing layer configurations and hyperparameter values used in CNNs can either create or resolve the issue of overfitting and affect final accuracy levels of object recognition systems. The main contribution of this thesis research is an evaluation of various approaches in structuring and training deep learning neural network object recognition algorithms. The experiment discovers what patterns exist in the hyperparameters and layering designs to achieve high performance under limited computational resources.
Advisor(s) or Committee Chair
Qi Li, Ph.D.
Artificial Intelligence and Robotics | Computer Sciences
Davis, Lillian, "Object Recognition with Deep Neural Networks in Low-End Systems" (2023). Mahurin Honors College Capstone Experience/Thesis Projects. Paper 1016.