Abstract
This study employs a T5 regression model for color theme extraction, addressing the limitations of traditional methods, such as high subjectivity and low efficiency. Experimental results show that the model achieves a maximum accuracy of 0.416 and an average color difference of 28.546, generating harmonious and contextually appropriate color themes. Additionally, its performance across different color quantization sets is analyzed, offering new insights into the integration of AI and design industries, and paving the way for future applications in automated visual communication.
| Original language | English |
|---|---|
| Title of host publication | International Conference on Applied System Innovation, ICASI 2025 |
| Publisher | Institution of Engineering and Technology |
| Pages | 53-55 |
| Number of pages | 3 |
| Volume | 2025 |
| Edition | 15 |
| ISBN (Electronic) | 9781837242634, 9781837243143, 9781837243150, 9781837243167, 9781837243235, 9781837243341, 9781837243358, 9781837246847, 9781837246854, 9781837247004, 9781837247011, 9781837247028, 9781837247035, 9781837247042, 9781837247271 |
| DOIs | |
| Publication status | Published - 2025 |
| Event | 2025 International Conference on Applied System Innovation, ICASI 2025 - Tokyo, Japan Duration: 2025 Apr 22 → 2025 Apr 25 |
Conference
| Conference | 2025 International Conference on Applied System Innovation, ICASI 2025 |
|---|---|
| Country/Territory | Japan |
| City | Tokyo |
| Period | 2025/04/22 → 2025/04/25 |
Keywords
- COLOR QUANTIZATION
- COLOR THEME
- SUPERPIXEL
- T5 MODEL
- TRANSFORMER
ASJC Scopus subject areas
- General Engineering
Fingerprint
Dive into the research topics of 'COLOR THEME EXTRACTION BASED ON THE T5 REGRESSION MODEL'. Together they form a unique fingerprint.Cite this
- APA
- Standard
- Harvard
- Vancouver
- Author
- BIBTEX
- RIS