To access this work you must either be on the Smith College campus OR have valid Smith login credentials.

On Campus users: To access this work if you are on campus please Select the Download button.

Off Campus users: To access this work from off campus, please select the Off-Campus button and enter your Smith username and password when prompted.

Non-Smith users: You may request this item through Interlibrary Loan at your own library.

Publication Date

2025-5

First Advisor

Shinyoung Cho

Document Type

Honors Project

Degree Name

Bachelor of Arts

Department

Computer Science

Keywords

traffic surveillance videos, memorability, multi-target multi-camera tracking, privacy preservation, visual attention

Abstract

Traffic surveillance plays an important role in modern society by enabling traffic flow optimization, accident detection, and urban planning. However, it also raises significant privacy concerns, as vehicle features captured in public footage can inadvertently reveal personal information. While prior approaches have focused primarily on preventing algorithmic identification, they often overlook privacy risks introduced by human perception. This research addresses that gap by examining privacy through the lens of visual memorability — the likelihood that a human casually observing surveillance footage retains memory of specific vehicles. This paper makes three core contributions: (1) It proposes a novel privacy perspective by analyzing human memory as a vector for passive information exposure; (2) It develops a complete pipeline to estimate frame-level vehicle memorability based on attention modeling and object-level features; and (3) It explores visual style transformation as a strategy to reduce memorability while preserving tracking utility. By analyzing which visual attributes contribute to a vehicle’s memorability using an attention-based pipeline, this work finds that simple visual features, such as size, position, and color, significantly affect whether a vehicle is remembered, offering insights for designing privacy-aware surveillance systems.

Rights

©2025 Yuxin Gao. Access limited to the Smith College community and other researchers while on campus. Smith College community members also may access from off-campus using a Smith College log-in. Other off-campus researchers may request a copy through Interlibrary Loan for personal use.

Language

English

Comments

v, 35 pages: color illustrations. Includes bibliographical references (pages 31-35).

Download

Smith Only:
Off Campus Download

Share

COinS