Publications

2025

1. 

   DentAll: A unified approach for tooth instance segmentation and missing tooth detection and localization using contextual spatial attention, positional encoding, and missing tooth location predictor

   Shaily Bajpai, Hamza Shafiq, and Bumshik Lee

   Under Review at Engineering Apllication of Artificial Intelligence (EAAI), 2025

2. 

   Colorformer: A Novel Colorization Method Based on a Transformer

   Hamza Shafiq, Truong Nguyen, and Bumshik Lee

   In Revision (Neurocomputing), 2025

2024

1. Electronics

   Transforming color: A novel image colorization method

   Hamza Shafiq, and Bumshik Lee

   Electronics, 2024

   DOI

2023

1. IEEE Access

   Image Colorization Using Color-Features and Adversarial Learning

   Hamza Shafiq, and Bumshik Lee

   IEEE Access, 2023

   DOI

Conference Papers

• PASR: Parallel Attention Image Super Resolution - 한국통신학회 학술대회논문집 (2024)
• An image colorization method using a transformer - 한국통신학회 학술대회논문집 (2023)
• TransVivid: Reviving Visuals with Transformer-based Colorization - 인공지능신호처리 학술대회 (2023)
• ColorGAN: Generative Adversarial Network based Image Colorization - Proceedings of KIIS Autumn Conference (2022)

Patents



트랜스포머 블록을 포함하는 GAN 기반의 모델을 이용하여 흑백 이미지를 컬러링하기 위한 전자 장치 및 그 동작 방법
Electronic device and method for colorizing black and white images using a GAN-based model including transformer blocks
Patent No: KR 10 2774306
Publication Date: 2025-02-25
Inventors: Hamza Shafiq et al.

This patent describes a novel approach to image colorization using a generative adversarial network (GAN) architecture enhanced with transformer blocks. The method achieves superior colorization results by leveraging the attention mechanisms of transformers with the generative capabilities of GANs.