Kang Cheol Kim

Focused on applied research in multimodal foundation models, including vision-language pretraining, image-text retrieval, and visual question answering. • Developed multimodal models based on open-source frameworks, adapting core modules and conducting large-scale training on public datasets (100M image-text pairs). • Handled end-to-end experimentation: dataset preparation, model adaptation, training pipeline implementation, and result validation. • Led feasibility studies on text-to-image generation with diffusion models for B2B scenarios.

Conducted research on deep learning-based medical image analysis