C

Clip Vit Base Patch32 Ko

Developed by Bingsu
Korean CLIP model trained via knowledge distillation, supporting Korean-English bilingual image-text matching tasks
Downloads 3,147
Release Time : 9/16/2022

Model Overview

This is a Korean version of the CLIP model based on the ViT-Base-Patch32 architecture, trained using knowledge distillation methods, specifically designed for Korean and English cross-modal retrieval tasks.

Model Features

Korean optimization
Specifically optimized for Korean, trained using Korean-English parallel corpus from AIHUB platform
Knowledge distillation training
Uses knowledge distillation to transfer learning from the original CLIP model
Bilingual support
Supports both Korean and English text inputs

Model Capabilities

Zero-shot image classification
Image-text matching
Cross-modal retrieval

Use Cases

Image classification
Animal recognition
Identify animal types in images
Can accurately distinguish common animals like cats and dogs
Content moderation
Inappropriate content detection
Detect if images contain inappropriate content
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase