Improving OCEAN Personality Recognition Using Vision Transformers and 2D Patch Partitioning 


Vol. 14,  No. 7, pp. 542-547, Jul.  2025
https://doi.org/10.3745/TKIPS.2025.14.7.542


PDF
  Abstract

As personalized services that reflect individuals’ personality traits and tendencies gain popularity, research in this area is rapidly advancing. In particular, the OCEAN model plays a significant role in personality recognition assessment. Generally, fine-tuning input images before training visual models can improve performance, with the 2D Patch Partition mechanism being one of the commonly used methods. While the conventional 2D patch partition method directly divides an image into four equal parts, the technique proposed in this study involves dividing the input image into smaller patches and then aggregating pixels at the same positions to generate four new images. The 1-MAE performance of these two methods was compared using the VST (Video Swin Transformer) and ViVit (Video Vision Transformer) models. By evaluating the 1-MAE performance of both the conventional 2D Patch Partition mechanism and our proposed improved version, we found that our method enhanced the 1-MAE performance of both models by approximately 0.001.

  Statistics
  Cite this article

[IEEE Style]

X. Qiu and B. Kim, "Improving OCEAN Personality Recognition Using Vision Transformers and 2D Patch Partitioning," The Transactions of the Korea Information Processing Society, vol. 14, no. 7, pp. 542-547, 2025. DOI: https://doi.org/10.3745/TKIPS.2025.14.7.542.

[ACM Style]

Xu Qiu and Bongjae Kim. 2025. Improving OCEAN Personality Recognition Using Vision Transformers and 2D Patch Partitioning. The Transactions of the Korea Information Processing Society, 14, 7, (2025), 542-547. DOI: https://doi.org/10.3745/TKIPS.2025.14.7.542.

Forms

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)

Old Journals

Indexing

All publications of TKIPS is indexed in DOI, EBSCO, Google Scholar, Crossref, and CrossCheck.

TKIPS is also selected as the Journal for Accreditation by NRF (National Research Foundation of Korea).

Related Journals