A5081467843- Topic Modeling
- Flood Risk Assessment and Management
- Natural Language Processing Techniques
- Tropical and Extratropical Cyclones Research
- Multimodal Machine Learning Applications
- Coastal and Marine Dynamics
- Vehicle License Plate Recognition
- Advanced Mathematical Modeling in Engineering
- Domain Adaptation and Few-Shot Learning
- Advanced Text Analysis Techniques
- Hydrology and Watershed Management Studies
- Language and cultural evolution
- Computational and Text Analysis Methods
- Generative Adversarial Networks and Image Synthesis
- Coastal wetland ecosystem dynamics
- Heat and Mass Transfer in Porous Media
- Soil and Unsaturated Flow
- Food Supply Chain Traceability
- Advanced Image and Video Retrieval Techniques
- Human-Animal Interaction Studies
- Aeolian processes and effects
- Handwritten Text Recognition Techniques
- Human Pose and Action Recognition
- Industrial Vision Systems and Defect Detection
- Groundwater flow and contamination studies
Amsterdam University of the Arts
2023-2024
University of Amsterdam
2023-2024
Ho Chi Minh City University of Technology
2017-2023
Vietnam National University Ho Chi Minh City
2017-2022
Ho Chi Minh City International University
2022
Norwegian University of Science and Technology
2022
VinUniversity
2021
Recently, Transformer-based models have been proven effective in the abstractive summarization task by creating fluent and informative summaries. Nevertheless, these still suffer from short-range dependency problem, causing them to produce summaries that miss key points of document. In this paper, we attempt address issue introducing a neural topic model empowered with normalizing flow capture global semantics document, which are then integrated into model. addition, avoid overwhelming...
The present study focuses on the long-term multi-year evolution of shoreline position Nha Trang sandy beach. To this end an empirical model which is a combination longshore and cross-shore models, used. beach morphology driven by tropical wave climate dominated seasonal variations winter monsoon intra-seasonal pulses. combined accounts for evolution, primarily attributed to dynamics but fails represent accretion that occurs during height summer under low energy conditions. reason in single...
This paper introduces our solution for Track 2 in AI City Challenge 2022. The task is Tracked-Vehicle Retrieval by Natural Language Descriptions with a real-world dataset of various scenarios and cameras. We mainly focus on developing robust natural language-based vehicle retrieval system to address the domain bias problem due unseen multi-view multi-camera tracks. Specifically, we apply CLIP [16] effectively extract both visual textual representations contrastive representation learning....
Learned sparse retrieval (LSR) is a family of neural methods that transform queries and documents into weight vectors aligned with vocabulary. While LSR approaches like Splade work well for short passages, it unclear how they handle longer documents. We investigate existing aggregation adapting to find proximal scoring crucial long To leverage this property, we proposed two adaptations the Sequential Dependence Model (SDM) LSR: ExactSDM SoftSDM. assumes only exact query term dependence,...
The paper deals with an Integrated Urban Flood Management Strategy for Ho Chi Minh City to adapt uncertainties of both natural and anthropogenic impacts. Observed data analysis hydraulic simulations were utilised prove the significance such uncertainties. Hydraulic proved that conventional solutions (e.g. storm sewers, dikes tide gates) may not be effective cope rainfall intensity increase land subsidence.Beyond constructional intervention will still required improve, soonest as possible,...
In this paper, we propose a system for Multi-Camera Multi-Target (MCMT) Vehicle Tracking in Track 1 of AI City Challenge 2022. There are many technical difficulties to the MCMT problem such as common lack labeled data real scenarios, distortion vehicle detailed appearances recording, and ambiguity between highly similar vehicles. Taking those into account, develop 3-component that exploits behavior, leverages synthetic multiple augmentation techniques, enforces contextual constraints....
Dynamic topic models track the evolution of topics in sequential documents, which have derived various applications like trend analysis and opinion mining. However, existing suffer from repetitive unassociated issues, failing to reveal hindering further applications. To address these we break tradition simply chaining work propose a novel neural \modelfullname. We introduce new evolution-tracking contrastive learning method that builds similarity relations among dynamic topics. This not only...
Learned Sparse Retrieval (LSR) techniques use neural machinery to represent queries and documents as learned bags of words. In contrast with other retrieval techniques, such generative dense retrieval, LSR has been shown be a remarkably robust, transferable, efficient family methods for retrieving high-quality search results. This half-day tutorial aims provide an extensive overview LSR, ranging from its fundamentals the latest emerging techniques. By end tutorial, attendees will familiar...
The Mekong Delta has the world’s third-largest surface area. It plays an indisputable role in economy and livelihoods of Vietnam Cambodia, with repercussions at regional global scales. During recent decades, Vietnamese part underwent profound human interventions (construction dykes multi-channel networks), which modified hydrodynamic regime, especially cycles field submersion. In this study, we first applied a full 2D numerical hydraulic model, TELEMAC-2D, to examine effects complex channel...
Solute diffusion is a key process in many fields like for example material science or environmental engineering. Diffusion mechanism porous media often described by Fick’s law. However, we could not use this law nonstandard behaviors occurring cases of heterogeneous media. The conception double-porosity medium can be applied to class such characterized two distinct pore sizes: macro-porosity domain and micro-porosity domain, respectively, having the contrasted hydraulic properties. This...
Review on the Evolution of Low-
Abstract The Mekong Delta has the world's third largest surface area. It plays an indisputable role in economy and livelihoods of Vietnam, Cambodia, with repercussions at regional global scales. During recent decades, Vietnamese part underwent profound human interventions (construction dykes multi-channel networks), which modified hydrodynamic regime, especially cycles field submersion. In this study, we first applied a full 2D numerical hydraulic model, TELEMAC-2D, to examine effects...
Recently, Transformer-based models have been proven effective in the abstractive summarization task by creating fluent and informative summaries. Nevertheless, these still suffer from short-range dependency problem, causing them to produce summaries that miss key points of document. In this paper, we attempt address issue introducing a neural topic model empowered with normalizing flow capture global semantics document, which are then integrated into model. addition, avoid overwhelming...