Other

AI Institute’s Publications

 

Semantic Alignment in Hyperbolic Space for
Open-Vocabulary Semantic Segmentation
Hoang M. Truong Hai Nguyen-Truong Dang Huynh*
CVPRW 2026, Pixel-level Video Understanding in the Wild (PVUW)
[Paper][Code]
Text-prior-guided super-resolution for French license plate recognition via diversity-augmented OCR ensemble
Phuc L. L. Nguyen, Dang Huynh*
ICIPW 2026, Extreme In-The-Wild License Plate Super-Resolution (XLPSR) Challenge
[Paper][Code]

The Agentic AI Accelerator programme, launched on 1 August 2026 under the Singapore Government’s Enterprise Compute Initiative (ECI) and led by Microsoft and Digital Industry Singapore, supports Singapore-based companies in adopting agentic AI autonomous systems. Photo: Singapore Economic Development Board (EDB) / Facebook.

Photo: Singapore Economic Development Board (EDB) / Facebook

Southeast Asia and AI Adoption: The Return of Cold War Strategies
Tobias Burgers, Tran Nguyen
FULCRUM, ISEAS – Yusof Ishak Institute, 31 August 2026
[Commentary]

Overview of Vendi information gain (VIG) for active learning: a dropout neural network samples labels for an unlabeled camera-trap image, is retrained on these sampled labels to predict the entire image pool, and VIG selects the image that most reduces uncertainty across the pool. Figure: Quan Nguyen and Adji Bousso Dieng / arXiv.

Figure: Quan Nguyen & Adji Bousso Dieng / arXiv

Vendi information gain for active learning and its application to ecology
Quan Nguyen, Adji Bousso Dieng
Methods in Ecology and Evolution, 24 October 2026
[Paper]

Overview of the iPerturb framework: a cell-type-specific template gene regulatory network (GRN) is built from public gene-regulation databases, the kinetic parameters of each edge are learnt from CRISPRi Perturb-seq data, and gene expression changes for unseen perturbations are predicted by multi-hop message passing on the trained GRN. Figure: Khang Ta & Linh Huynh / CIBB 2026.

Figure: Khang Ta & Linh Huynh / CIBB 2026

iPerturb: An interpretable framework for predicting gene expression changes under perturbations
Khang Ta, Linh Huynh*
CIBB 2026, 21st International Conference on Computational Intelligence Methods for Bioinformatics and Biostatistics, Sapienza University of Rome, Italy, 2–4 September 2026 (Oral presentation)