��Identifying Fashion Trends Utilizing Color Analysis

Hehjun Lim

Albert Dorman Honors College

​

Adam Spryszynski

Department of Informatics

Problem Statement and Goals

• Problem Statement:

​

​

What are some fashion trends that are prominent today?

​

• Project Goals:

​

Utilizing images of clothing online to provide insight to consumer behavior, cultural influences, and market dynamics.

​

​

Need and Significance

• Fashion designers, retailers, and industry professionals can learn about the preferences and choices of consumers to cater their business and services to them.

​

• Individuals can quickly identify popular trends as inspiration for their personal style.

​

• Finding a technology solution will simplify the process fashion trend analysis.

Background

• Color plays a significant role in fashion product development and design.

​

• Companies have taken advantage of technology to quickly spread fashion trends and market their products.

​

​

• Previous Research:
• Distinguishing clothing items in images through machine learning.
• Retrieving and grouping items of clothing that have similar characteristics

Research Approach and Methods

Removal of background for all images

Perform color analysis and collect data into a dataframe using Python and Jupyter Notebook.

Analyze recorded data and create visualizations.

All Colors

&

Original

All Colors

& Processed

Dominant Colors

&

Original

Dominant Colors

&

Processed

Sample of Results

Innovation, Discussion and Future Work

• Bridging the gap between data-driven analysis and the ever-evolving world of fashion.

​

​

• Diversity and creativity of fashion is exemplified by utilizing photos created and posted by individuals from various backgrounds.

​

​

• Promotes inclusivity by demonstrating that professionally taken photos are not necessary for determining future clothing trends or patterns.

​

​

​

​

Future Works

​

• Scraping images from multiple social media platforms to perform fashion trend analysis specifically on images posted by content creators.

​

• Innovate methodologies on studying and analyzing streetwear fashion beyond color analysis.

​

• Focus on creating more relevant fashion databases to enhance applicability of future research.

Acknowledgment

I would like to thank Adam Spryszynski for supervising this research project.

​

I would also like to thank the Albert Dorman Honors College for sponsoring this project.

​