Trung-Nghia Le

Welcome to my homepage! I am a senior researcher and lecturer at University of Science's Software Engineering Laboratory and Faculty of Information Technology.

I received my B.Sc (Honors Program), M.Sc, Ph.D in Computer of Science at University of Science, John Von Neumann Institute, and National Institute of Informatics in 2012, 2014, and 2018 respectively.

I currently lead the Multimedia Synergy (MS) research group to advance research at the crossroads of computer vision, multimodal intelligence, and generative AI. We study how machines can see, imagine, and reason about the world through synergistic integration of vision, language, and creativity. Our mission is to create systems that not only perceive reality but also generate meaningful and trustworthy multimedia content, bridging analysis, synthesis, and understanding.

🔥 Research topics of interest:

Visual Perception and Understanding
Multimodal and Context-Aware Learning
Generative and Creative AI
Trustworthy and Forensic AI
Collaborative & Human-Centered AI Systems
Health Science

🙋 Prospective Students:

I am looking for motivated undergraduate/graduate students. Please send an email to ltnghia at fit.hcmus.edu.vn to discuss your research interests.

News

Some challenge and demo papers are accepted to ACM MM, 2025.
OpenEventsV1 dataset is accepted to ACM MM, 2025.
A paper about semi-supervised segmentation is accepted to MICCAI Workshop, 2025.
A paper about information retrieval is accepted to CBMI, 2025.
A paper about 3D retrieval is accepted to MAPR, 2025.
An article about camouflage segmentation is accepted to Journal of Neural Computing & Applications, 2025.
Three papers (image generation and application) are accepted to ICCCI, 2025.
We organize CBMI Special Session "Multimedia AI in Modern CB Retrieval: Challenges and Applications", 2025.
We organize 1st Workshop on Ambiguous Object Analysis in Computer Vision (AOCV), ACM MM, 2025.
We organize 2nd Workshop on Large Vision-Language Model Learning and Applications (LAVA), ACM MM, 2025.
We organize Event-Enriched Image Analysis (EVENTA) Challenge, ACM MM, 2025.
We organize Advancing Vision-Language AI for ENT Endoscopy Analysis (ENTRep) Challenge, ACM MM, 2025.
A paper about camouflage segmentation is accepted to WACV, 2025..
Some papers are accepted to SoICT, 2024.
Two papers are accepted to ACCV, 2024.
Two poster papers are accepted to ISMAR, 2024.
My student Quang-Binh Nguyen joins VinAI Residency, 2024
We organize 1st Workshop on Large Vision-Language Model Learning and Applications (LAVA), ACCV, 2024.
We achieve top-1 result in the CVPR SyntaGen Challenge, 2024.
Two papers are accepted to CHI (Late Breaking Work track), 2024.
An article is accepted to the Journal of Imaging Informatics in Medicine, 2024.
A paper about few-shot instance segmentation is accepted to AAAI, 2024 (Oral presentation).
A paper about self-supervised learning is accepted to MMM, 2024 (Oral presentation).
We organize Medico track of MediaEval, MMM, 2024.
Two papers are accepted to PSIVT, 2023 (one achieved Best Paper Award)
Our virtual try-on paper is nominated for Best Poster at ISMAR 2023.
Two poster papers are accepted to ISMAR, 2023.
We organize two tracks of ANIMAR Challenge at SHREC 2023.
Two papers (multimedia security and adversarial example) are accepted to WACV, 2023.
A paper about adversarial example is accepted to WIFS, 2022.
A poster paper about public speaking simulator is accepted to ISMAR, 2022.
An article about aquatic animal segmentation will be presented at CV4Animal Workshop, CVPR 2022.
An article about video object segmentation is accepted to MVA journal, 2022.
An article about deepfake is accepted as a book chapter of Frontiers in Fake Media Generation and Detection, 2021.
An article about camouflaged instance segmentation is accepted to IEEE Trans. on Image Processing (TIP, 2021).
An article about face analysis is accepted to Journal of Imaging, 2021.
A paper about deepfake detection is accepted to ICCV 2021 (Acceptance rate 25.9%).
One CREST AIP proposal is funded (June 2021).
A paper about adversarial examples is accepted to Workshop of Media Forensics, CVPR 2021.
One JSPS KAKENHI proposal is funded (March 2021).
An article about camouflaged object segmentation is accepted to IEEE Access, 2021.
Call for papers: Special Issue "Advanced Artificial Intelligence in Mixed Reality Applications" of Journal of Electronics
Two papers (video object segmentation and camouflaged instance segmentation) are accepted to AAAI 2021, demonstration program.
One JSPS KAKENHI proposal is funded (September 2020).
One CREST AIP proposal is funded (May 2020).
A paper about text-to-image is accepted to ACM Multimedia 2020, demo track.
A paper about object detection is accepted to Intelligent Vehicles Symposium (IV) 2020.
Rank 4 in Densely Annotated VIdeo Segmentation (DAVIS) Challenge (Semi-Supervised Segmentation Track), CVPR 2020.
Rank 10 in AI City Challenge (Track 1), CVPR 2020.
Rank 5 in AI City Challenge (Track 4), CVPR 2020.
A paper about object detection is accepted to WACV 2020.
Rank 25 in AI City Challenge (Track 2), CVPR 2019 (out of 84 team submissions).
Rank 8 in AI City Challenge (Track 3), CVPR 2019.
Rank 3 in Densely Annotated VIdeo Segmentation (DAVIS) Challenge (Semi-Supervised Segmentation Track), CVPR 2019.
An article about camouflaged object segmentation is accepted to CVIU journal, 2019.
A paper about salient object segmentation is accepted to WACV 2019.
An article about salient object segmentation is accepted to IEEE Trans. on Image Processing (TIP, 2018).
Rank 6 in Densely Annotated VIdeo Segmentation (DAVIS) Challenge (Semi-Supervised Segmentation Track), CVPR 2018.
A paper about style transfer is accepted to WACV 2018.
Rank 3 in Densely Annotated VIdeo Segmentation (DAVIS) Challenge (Semi-Supervised Segmentation Track), CVPR 2017.
A paper about salient object segmentation is accepted to BMVC 2017.
A paper about salient object segmentation is accepted to PSIVT 2015.