Searcharxiv⌕ Search

arXiv subjects

Dung Nguyen

Publications and source records attributed to Dung Nguyen.

65 records · Page 4Linked to original sources
arXiv2020

Joint Deep Cross-Domain Transfer Learning for Emotion Recognition

Deep learning has been applied to achieve significant progress in emotion recognition. Despite such substantial progress, existing approaches are still hindered by insufficient training data, and the resulting models do not generalize well under mismatched conditions. To address this challenge, we propose a learning strategy which jointly transfers the knowledge learned from rich datasets to source-poor datasets. Our method is also able to learn cross-domain features which lead to improved recognition performance. To demonstrate the robustness of our proposed framework, we conducted experiments on three benchmark emotion datasets including eNTERFACE, SAVEE, and EMODB. Experimental results show that the proposed method surpassed state-of-the-art transfer learning schemes by a significant margin.

cs.CV↗
arXiv2019

Language Support for Adaptation: Intent-Driven Programming in FAST

Historically, programming language semantics has focused on assigning a precise mathematical meaning to programs. That meaning is a function from the program's input domain to its output domain determined solely by its syntactic structure. Such a semantics, fosters the development of portable applications which are oblivious to the performance characteristics and limitations (such as a maximum memory footprint) of particular hardware and software platforms. This paper introduces the idea of intent-driven programming where the meaning of a program additionally depends on an accompanying intent specification expressing how the ordinary program meaning is dynamically modified during execution to satisfy additional properties expressed by the intent. These include both intensional properties---e.g., resource usage---and extensional properties---e.g., accuracy of the computed answer. To demonstrate the intent-driven programming model's value, this paper presents a general-purpose intent-driven programming language---called FAST---implemented as an extension of Swift. FAST consists of an intent compiler, a profiler, a general controller interface and a runtime module which supports interoperation with legacy C/C++ codes. Compared to existing frameworks for adaptive computing, \FAST{} supports dynamic adaptation to changes both in the operating environment and in the intent itself, and enables the mixing of procedural control and control based on feedback and optimization.

cs.PL↗
arXiv2018

Meta Transfer Learning for Facial Emotion Recognition

The use of deep learning techniques for automatic facial expression recognition has recently attracted great interest but developed models are still unable to generalize well due to the lack of large emotion datasets for deep learning. To overcome this problem, in this paper, we propose utilizing a novel transfer learning approach relying on PathNet and investigate how knowledge can be accumulated within a given dataset and how the knowledge captured from one emotion dataset can be transferred into another in order to improve the overall performance. To evaluate the robustness of our system, we have conducted various sets of experiments on two emotion datasets: SAVEE and eNTERFACE. The experimental results demonstrate that our proposed system leads to improvement in performance of emotion recognition and performs significantly better than the recent state-of-the-art schemes adopting fine-\ tuning/pre-trained approaches.

cs.CV↗
arXiv2015

Join Processing for Graph Patterns: An Old Dog with New Tricks

Join optimization has been dominated by Selinger-style, pairwise optimizers for decades. But, Selinger-style algorithms are asymptotically suboptimal for applications in graphic analytics. This suboptimality is one of the reasons that many have advocated supplementing relational engines with specialized graph processing engines. Recently, new join algorithms have been discovered that achieve optimal worst-case run times for any join or even so-called beyond worst-case (or instance optimal) run time guarantees for specialized classes of joins. These new algorithms match or improve on those used in specialized graph-processing systems. This paper asks can these new join algorithms allow relational engines to close the performance gap with graph engines? We examine this question for graph-pattern queries or join queries. We find that classical relational databases like Postgres and MonetDB or newer graph databases/stores like Virtuoso and Neo4j may be orders of magnitude slower than these new approaches compared to a fully featured RDBMS, LogicBlox, using these new ideas. Our results demonstrate that an RDBMS with such new algorithms can perform as well as specialized engines like GraphLab -- while retaining a high-level interface. We hope this adds to the ongoing debate of the role of graph accelerators, new graph systems, and relational systems in modern workloads.

cs.DB↗
arXiv2011

Characteristic numbers of elliptic space curves

We solve the problem of characteristic numbers of elliptic curves in any dimensional projective space The answers are given in the form of effective recursions. Many numerical examples are provided. A C++ program implementing all the recursions is available upon request.

math.AG↗
arXiv2011

Characteristic numbers of rational cuspidal space curves

We solve the problem of computing characteristic numbers of rational space curves with a cusp, where there may or may not be a condition on the node. The solution is given in the form of effective recursions. We give explicit formulas when the dimension of the ambient projective space is at most 5. Many numerical examples are provided. A C++ program implementing most of the recursions is available upon request

math.AG↗
arXiv2011

Characteristic numbers of ellitpic space curves with fixed j-invariant

We solve the problem of counting elliptic curves with fixed j-invariant in projective space with tangency conditions. This is equivalent to couting rational nodal curves with condition on the node of the image. The solution is given in the form of effective recursions. We give explicit formulas when the dimension of the ambient projective space is at most 5. Many numerical examples are provided. A C++ program implementing most of the recursions is available on the author's webpage (see \cite{dn}).

math.AG↗
arXiv2010

A Surprise in the Amplitude/Wilson Loop Duality

One of the many remarkable features of MHV scattering amplitudes is their conjectured equality to lightlike polygon Wilson loops, which apparently holds at all orders in perturbation theory as well as non-perturbatively. This duality is usually expressed in terms of purely four-dimensional quantities obtained by appropriate subtraction of the IR and UV divergences from amplitudes and Wilson loops respectively. In this paper we demonstrate, by explicit calculation, the completely unanticipated fact that the equality continues to hold at two loops through order epsilon in dimensional regularization for both the four-particle amplitude and the (parity-even part of the) five-particle amplitude.

hep-th↗
arXiv2009

The Tree Formula for MHV Graviton Amplitudes

We present and prove a formula for the MHV scattering amplitude of n gravitons at tree level. Some of the more interesting features of the formula, which set it apart as being significantly different from many more familiar formulas, include the absence of any vestigial reference to a cyclic ordering of the gravitons--making it in a sense a truly gravitational formula, rather than a recycled Yang-Mills result, and the fact that it simultaneously manifests both S_{n-2} symmetry as well as large-z behavior that is O(1/z^2) term-by-term, without relying on delicate cancellations. The formula is seemingly related to others by an enormous simplification provided by O(n^n) iterated Schouten identities, but our proof relies on a complex analysis argument rather than such a brute force manipulation. We find that the formula has a very simple link representation in twistor space, where cancellations that are non-obvious in physical space become manifest.

hep-th↗
arXiv2007

New Dual Conformally Invariant Off-Shell Integrals

Evidence has recently emerged for a hidden symmetry of scattering amplitudes in N=4 super Yang-Mills theory called dual conformal symmetry. At weak coupling the presence of this symmetry has been observed through five loops, while at strong coupling the symmetry has been shown to have a natural interpretation in terms of a T-dualized AdS_5. In this paper we study dual conformally invariant off-shell four-point Feynman diagrams. We classify all such diagrams through four loops and evaluate 10 new off-shell integrals in terms of Mellin-Barnes representations, also finding explicit expressions for their infrared singularities.

hep-th↗