Duc-Tuan Luu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Duc-Tuan Luu has a total of 48 co-authors. Here are the top co-authors by number of shared publications:
- Vinh-Tiep Nguyen
University of Information Technology
6 shared publications - Duc-Tien Dang-Nguyen
University of Bergen, Bergen, Norway
4 shared publications - Minh-Triet Tran
University of Science
3 shared publications - Anh-Khoa Nguyen Vu
University of Information Technology
2 shared publications - Dang Van Thin
University of Information Technology, VNU-HCM
2 shared publications - Duc Minh Vo
The University of Tokyo
2 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 17599-17605
LAVA Grand Challenge 2026: Grounded Multi-Page Document Question Answering in Japanese and Vietnamese
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 17599-17605
Open AccessLAVA Grand Challenge 2026: Grounded Multi-Page Document Question Answering in Japanese and Vietnamese
- 2 September 2026
- Published by Springer Nature
- p. 81-95
RIOT: Robust Incremental Few-Shot Instance Segmentation via Synthetic Feature Generation with Optimal Transport
- 2 September 2026
- Published by Springer Nature
- p. 81-95
RIOT: Robust Incremental Few-Shot Instance Segmentation via Synthetic Feature Generation with Optimal Transport
- 15 August 2026
- Published by Springer Nature
- p. 229-240
OTGen-FSIS: Optimal Transport–Driven Feature Generation for Few-Shot Instance Segmentation
- 15 August 2026
- Published by Springer Nature
- p. 229-240
OTGen-FSIS: Optimal Transport–Driven Feature Generation for Few-Shot Instance Segmentation
- 27 June 2026
- Published by Springer Nature
- p. 279-293
From Slides to Exams: A Multi-agent Human-AI System for Collaborative Assessment Design
- 27 June 2026
- Published by Springer Nature
- p. 279-293
From Slides to Exams: A Multi-agent Human-AI System for Collaborative Assessment Design
- Open Access
- 16 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 101-106
VitaChronicle 3.0: Enhancing Interactive Lifelog Search with OCR and Scene Text Recognition (STR)
Retrieving specific moments from continuous lifelog collections is a complex challenge, particularly in highly competitive, time-constrained environments like the Lifelog Search Challenge (LSC). While CLIP-based vision-language models excel at general semantic matching, they frequently fail to capture explicit ambient text—such as brand names, receipts, and signage—which are critical anchors for identifying exact events. To address this challenge for LSC’26, we introduce VitaChronicle 3.0, which marks a strategic shift from a dual-mode interface to a consolidated and unified system that prioritizes high information density and granular control. This version includes a hybrid retrieval pipeline that fuses dense visual embeddings with a robust sparse index for Scene Text Recognition and optical character recognition, enabling precise filtering of discriminative cues. Additionally, the effective parameterized query design has been retained to maintain the user-centric experience, reduce cognitive load, and accelerate visual verification for complex multimodal lifelog exploration. By explicitly leveraging scene-text semantics within a streamlined user interface, VitaChronicle 3.0 accelerates visual verification and submission times for complex lifelog retrieval tasks.
- 16 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 101-106
Open AccessVitaChronicle 3.0: Enhancing Interactive Lifelog Search with OCR and Scene Text Recognition (STR)
Retrieving specific moments from continuous lifelog collections is a complex challenge, particularly in highly competitive, time-constrained environments like the Lifelog Search Challenge (LSC). While CLIP-based vision-language models excel at general semantic matching, they frequently fail to capture explicit ambient text—such as brand names, receipts, and signage—which are critical anchors for identifying exact events. To address this challenge for LSC’26, we introduce VitaChronicle 3.0, which marks a strategic shift from a dual-mode interface to a consolidated and unified system that prioritizes high information density and granular control. This version includes a hybrid retrieval pipeline that fuses dense visual embeddings with a robust sparse index for Scene Text Recognition and optical character recognition, enabling precise filtering of discriminative cues. Additionally, the effective parameterized query design has been retained to maintain the user-centric experience, reduce cognitive load, and accelerate visual verification for complex multimodal lifelog exploration. By explicitly leveraging scene-text semantics within a streamlined user interface, VitaChronicle 3.0 accelerates visual verification and submission times for complex lifelog retrieval tasks.
- 7 February 2026
- Published by Springer Nature
- p. 143-149
Past Forward: Exif Revisited
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 14041-14047
Ægis: AI-Enhanced OSINT for Multimedia Verification
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 14041-14047
Open AccessÆgis: AI-Enhanced OSINT for Multimedia Verification
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 32 (4) , 45-57
STERR-GAN: Spatiotemporal Rerendering for Facial Video Restoration
Old facial films are valuable historical resources that vividly depict significant figures of the past. However, these films suffer from low quality and visual distortion due to outdated camera technology and poor preservation. Facial video restoration offers a solution, incorporating sub-tasks such as noise removal, colorization, and facial enhancement. A naive approach that restores each frame independently often results in inconsistency. In this paper, we introduce Spatio-temporal Re-rendering for Facial Video Restoration (STERR-GAN), which features a Temporal Recurrent Network to leverage adjacent frame information for maintaining temporal coherence, and a specialized GAN module to enhance facial details. Our experimental results demonstrate that STERR-GAN significantly outperforms existing methods in both quantitative and qualitative analyses. Additionally, to address the insufficiency of current datasets, we propose the Facial Video dataset for Antique Restoration (Face-VAR) to support further research in facial video restoration.
- 18 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 32 (4) , 45-57
STERR-GAN: Spatiotemporal Rerendering for Facial Video Restoration
Old facial films are valuable historical resources that vividly depict significant figures of the past. However, these films suffer from low quality and visual distortion due to outdated camera technology and poor preservation. Facial video restoration offers a solution, incorporating sub-tasks such as noise removal, colorization, and facial enhancement. A naive approach that restores each frame independently often results in inconsistency. In this paper, we introduce Spatio-temporal Re-rendering for Facial Video Restoration (STERR-GAN), which features a Temporal Recurrent Network to leverage adjacent frame information for maintaining temporal coherence, and a specialized GAN module to enhance facial details. Our experimental results demonstrate that STERR-GAN significantly outperforms existing methods in both quantitative and qualitative analyses. Additionally, to address the insufficiency of current datasets, we propose the Facial Video dataset for Antique Restoration (Face-VAR) to support further research in facial video restoration.
- Open Access
- 30 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 63-69
VitaChronicle 2.0: A Dual-Mode UX-Centered Approach to Lifelog Image Retrieval for Novice and Expert Users
- 30 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 63-69
Open AccessVitaChronicle 2.0: A Dual-Mode UX-Centered Approach to Lifelog Image Retrieval for Novice and Expert Users
- 26 April 2025
- Published by Springer Nature
- p. 80-91
KPI: Knowledge-Based Processing for Interactive Video Retrieval
- 26 April 2025
- Published by Springer Nature
- p. 80-91
KPI: Knowledge-Based Processing for Interactive Video Retrieval