Masters Theses

Keywords and Phrases

Computer Vision; Deep Learning; Object Detection


"Computer vision based on deep learning is an essential field that plays a significant role in object detection, image classification, semantic segmentation, instance segmentation, and other applications. However, these models face significant challenges in adverse conditions, such as small objects, low-resolution images, and edge deployment. These challenges limit the accuracy and efficiency of computer vision algorithms, making it difficult to obtain reliable results.

The primary objective of this thesis is to assess the performance of deep learning- based computer vision models in challenging conditions and provide viable solutions to overcome the obstacles. The study will specifically address three key challenges, namely, the detection of small objects, handling low-resolution images, and deployment of models at the edge.

To address the challenges of small objects and low-resolution images, we propose SPD-Conv. This new CNN building block eliminates strided convolution and pooling layers to improve the detection of small objects and reduce the loss of fine-grained information. To address the challenge of edge deployment, we propose YOGA, a lightweight object detection model that achieves high accuracy on low-end edge devices by using a two-phase feature learning pipeline with attention-based multi-scale feature fusion.

The proposed solutions are evaluated on COCO-val and COCO-testdev datasets and compared with state-of-the-art models, demonstrating their effectiveness in overcoming these challenging scenarios. This thesis places significant emphasis on the importance of reproducibility in research. All experiments are conducted using open-source tools and frameworks, and the code and models are made available to the research community. This ensures that the results are transparent, and others can easily reproduce and build upon the work presented in this thesis"--Abstract, p. iv


Luo, Tony T.

Committee Member(s)

Nadendla, V. Sriram Siddharth
Tripathy, Ardhendu S.


Computer Science

Degree Name

M.S. in Computer Science


Missouri University of Science and Technology

Publication Date

Spring 2023


x, 56 pages

Note about bibliography

Includes_bibliographical_references_(pages 51-53)


© 2023 Raja Sunkara, All Rights Reserved

Document Type

Thesis - Open Access

File Type




Thesis Number

T 12264

Electronic OCLC #