The World is Full of Clocks: Information Underneath Visualizations

Webcam images have been useful in many different areas, they recordhow a scene changes over large time scales.Between the changes, there may be clues to the time of day. Deep neural networks have been widely used nowadays, especially in the field of computer vision. But the experiment of predicting time from sight is rarely explored. In this paper, I try to use some deep neural network to 1) determine the area that the model thinks best represents a particular time for a particular scene. 2) find out what triggers the model to detect the change, human activities or natural scenes. 3)Information underneath those visualizations, especially for those wrong prediction images. vi

Author

• Shen, Jingzhou

Language

• en

Keyword

• Computer vision
• machine learning

Date created

• 2023

Type of Work

• Master's Thesis

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Computer Science

Degree

• M.S.

Advisor

• Aviv, Adam

Committee Member(s)

• Bulusu, Kartik
• Pless, Robert

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/3t945r514