VNU-UET Repository: No conditions. Results ordered -Date Deposited. 2024-03-28T21:00:24ZEPrintshttp://eprints.uet.vnu.edu.vn/images/sitelogo.pnghttps://eprints.uet.vnu.edu.vn/eprints/2020-12-10T02:44:46Z2020-12-10T02:44:46Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/4218This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/42182020-12-10T02:44:46ZKey phrase Generation for Vietnamese Administrative Documents: A Collaborative ApproachThi Thu Trang NguyenThi Hai Yen Vuongyenvth_57@vnu.edu.vnVan Lien Tran14020768@vnu.edu.vnLe Minh NguyenXuan Hieu Phanhieupx@vnu.edu.vn2019-06-20T22:49:54Z2019-06-20T22:49:54Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/3511This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/35112019-06-20T22:49:54ZQASA: Advanced Document Retriever for Open-Domain Question Answering by Learning to Rank Question-Aware Self-Attentive Document RepresentationsFor information consumers, being able to obtain a short and accurate answer for a query is one of the most desirable features. This motivation, along with the rise of deep learning, has led to a boom in open-domain Question Answering (QA) research. While the problem of machine comprehension has received multiple success with the help of large training corpora and the emergence of attention mechanism, the development of document retrieval in open-domain QA is lagged behind. In this work, we propose a novel encoding method for learning question-aware self-attentive document representations. By applying pair-wise ranking approach to these encodings, we build a Document Retriever, called QASA, which is then integrated with a machine reader to form a complete open-domain QA system. Our system is thoroughly evaluated using QUASAR-T dataset and shows surpassing results compared to other state-of-the-art methods.Minh Trang Nguyentrangnm@vnu.edu.vnVan Lien Tran14020768@vnu.edu.vnDuy Cat Cancatcd@vnu.edu.vnQuang Thuy Hathuyhq@vnu.edu.vnThi Ly Vutlvu@ntu.edu.sgEng-Siong ChngASESChng@ntu.edu.sg2019-06-04T14:52:07Z2019-06-04T14:52:07Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/3474This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/34742019-06-04T14:52:07ZQASA: Advanced Document Retriever for Open-Domain Question Answering by Learning to Rank Question-Aware Self-Attentive Document RepresentationsFor information consumers, being able to obtain a short and accurate answer for a query is one of the most desirable features. This motivation, along with the rise of deep learning, has led to a boom in open-domain Question Answering (QA) research. While the problem of machine comprehension has received multiple success with the help of large training corpora and the emergence of attention mechanism, the development of document retrieval in open-domain QA is lagged behind. In this work, we propose a novel encoding method for learning question-aware self-attentive document representations. By applying pair-wise ranking approach to these encodings, we build a Document Retriever, called QASA, which is then integrated with a machine reader to form a complete open-domain QA system. Our system is thoroughly evaluated using QUASAR-T dataset and shows surpassing results compared to other state-of-the-art methods.Thi Minh Trang NguyenVan Lien Tran14020768@vnu.edu.vnDuy Cat Cancatcd@vnu.edu.vnQuang Thuy Hathuyhq@vnu.edu.vnThi Ly VuEng Siong ChngASESChng@ntu.edu.sg2018-12-08T08:49:26Z2018-12-08T08:49:26Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/3195This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/31952018-12-08T08:49:26ZDepth camera based navigation algorithms for indoor mobile robotHow can we efficiently search an object in a room? This report introduces a method for a single indoor mobile robot to find a hidden item based on states of the room when the robot is moving. A 2D distribution, called cognitive map, is built during robot movements to boost the exploring time. It is known that in the filed of exploring algorithms, A∗ usually takes more time to reach the target than recent invented algorithms such as rapidly-exploring random trees (RRT) and probabilistic roadmap (PRM). However, by adapting the cognitive map as a cost map, the A∗ algorithm is significantly improved and surpasses the two algorithms in Scannet 3D dataset. We also introduce application of depth sensors and SLAM solvers on reconstructing the room and updating cognitive map. By running a virtual robot in Gazebo simulator, it is proved that our method can work well on synthetic environment and hence, is very promising to be worked on real-life environment.Van Lien TranViet Thang NguyenCong Hoang Quachhoangqc@vnu.edu.vnXuan Hieu Phanhieupx@vnu.edu.vnMinh Trien Phamtrienpm@vnu.edu.vn2017-11-05T13:09:51Z2017-11-06T02:29:11Zhttp://eprints.uet.vnu.edu.vn/eprints/id/eprint/2614This item is in the repository with the URL: http://eprints.uet.vnu.edu.vn/eprints/id/eprint/26142017-11-05T13:09:51ZReal-time Lane Marker Detection Using Template Matching with RGB-D CameraThis paper addresses the problem of lane
detection which is fundamental for self-driving vehicles.
Our approach exploits both colour and depth information
recorded by a single RGB-D camera to better deal with
negative factors such as lighting conditions and lane-like
objects. In the approach, the colour and depth images
are first converted to a half-binary format and a 2D
matrix of 3D points. Those representations are then used
as inputs of template matching and geometric feature
extraction processes to calculate a response map that
its values present the probability of pixels being lane
markers. To enhance the result, the principal component
analysis and lane model fitting techniques are employed
to refine the template and form lane surfaces. A number
of experiments have been conducted on both synthetic
and real datasets. The result shows that the proposed
approach can effectively eliminate the unwanted noise
to accurately detect lane markers in various scenarios.
With the hardware configuration of a popular laptop
computer, the program implementation operates at the
speed of 20 frames per second which is sufficient for
real-time autonomous driving applications.Cong Hoang Quachhoangqc@vnu.edu.vnManh Duong Phungduongpm@vnu.edu.vnMinh Trien Phamtrienpm@vnu.edu.vnHung Nguyen14020780@vnu.edu.vnThang Nguyen16020048@vnu.edu.vnVan Lien Tran14020768@vnu.edu.vn