DR-NTU (Data) (Nanyang Technological University)
DR-NTU (Data) (Nanyang Technological University)Not a member yet
1955 research outputs found
Sort by
Zoo-Lǎohǔ word set: A cross linguistic lexical priming word set for animacy judgements (English and Mandarin Chinese)
No full textAudio tokens selected and edited from BLIP Lab's Singapore Early Word List - Audio Recordings. Selection performed by Tong, Zhane C. in 2023, under the supervision of SJS. Original audio recordings conducted by Woon Fei Ting and Annabel Loh under the supervision of Suzy J Styles as part of the Talkathon study in 2022-4. Tokens in both languages were spoken by a female Singaporean early-parallel bilingual of Singapore English and Singapore Mandarin in her 20s. For full details of audio recording conditions, please consult the main repository.
Audio tokens are stereo wav files. Each token is a single word listed in the file name. This repository contains 86 English and 82 Mandarin tokens of words that are suitable for use with children under the age of 5 in Singapore. For English tokens, the name of the audio file is the word in English. To facilitate checking, handling and picture matching by our multilingual team, for the Mandarin tokens, the file names are given as English translation equivalents followed by the letter M. For Pinyin and Hanzi of the Mandarin words please consult the AudioFileList. Use the "tree" to view files in folder context.
The majority of these words were used in a semantic decision task (animate/inanimate), with thematic semantic priming (related/unrelated) within and across languages (language match/mismatch). The study was preregistered (https://doi.org/10.21979/N9/ERB6J8), and materials, data and code have been archived (https://doi.org/10.21979/N9/JXMRVM). The list of semantic priming pairs is included in this repository
Replication Data for: Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Vision
No full textWe present Q-Bench, a holistic benchmark crafted to systematically
evaluate potential abilities of MLLMs on three realms: low-level visual perception, low-level visual description, and overall visual quality assessment
Related Data for: Aperiodic Bragg Reflectors for Tunable High-Purity Structural Color Based on Phase Change Material
No full textData for all published figure
Replication Data for: Nonlinear finite-difference time-domain method for exciton-polaritons: Application to saltatory conduction in polariton neurons
No full textThis data set contains all necessary data to reproduce figures 1b, 3, 4, 5
Data appendix for: Holocene relative sea-level histories of tropical islands
No full text This is the data appendix for the PhD thesis:
Tan, F. (2023). Holocene relative sea-level histories of tropical islands.
The dataset contains nine excel files. Tables A6 to A10 are supporting tables for appendix A; tables B4 to B5 are are supporting tables for appendix B; tables C1 to C2 are are supporting tables for appendix C.</p
Related Data for: Cascaded FSO systems with optical reflecting surfaces
No full textMATLAB and Python source code the publication title: "Cascaded FSO systems with optical reflecting surfaces"
These code will produce the outage probability and Bit error rate plots for the above pape
Related Data for: Fingerprinting Deep Neural Networks - a DeepFool Approach
No full textThis dataset contains partial program source code and the data for analysis for the paper "Fingerprinting Deep Neural Networks - a DeepFool Approach
Replication Data for: MMLongBench-Doc: Benchmarking Long-context Document Understanding with Visualizations
No full textUnderstanding documents with rich layouts and multi-modal components is a long-standing and practical task. Recent Large Vision-Language Models (LVLMs) have made remarkable strides in various tasks, particularly in single-page document understanding (DU). However, their abilities on long-context DU remain an open problem. This work presents MMLongBench-Doc, a long-context, multi-modal benchmark comprising 1,062 expert-annotated questions. Distinct from previous datasets, it is constructed upon 130 lengthy PDF-formatted documents with an average of 49.4 pages and 20,971 textual tokens. Towards comprehensive evaluation, answers to these questions rely on pieces of evidence from (1) different sources (text, image, chart, table, and layout structure) and (2) various locations (i.e. page number). Moreover, 33.2% of the questions are cross-page questions requiring evidence across multiple pages. 22.8% of the questions are designed to be unanswerable for detecting potential hallucinations. Experiments on 14 LVLMs demonstrate that long-context DU greatly challenges current models. Notably, the best-performing model, GPT-4o, achieves an F1 score of only 42.7%, while the second-best, GPT-4V, scores 31.4%. Furthermore, 12 LVLMs (all except GPT-4o and GPT-4V) even present worse performance than their LLM counterparts which are fed with lossy-parsed OCR documents. These results validate the necessity of future research toward more capable long-context LVLMs
Related Data for: Light-evoked deformations in rod photoreceptors, pigment epithelium and subretinal space revealed by prolonged and multilayered optoretinography
No full textIt contains two example datasets for demonstrating our prolonged and multilayered optoretinography (ORG) method (https://doi.org/10.1038/s41467-024-49014-5). The corresponding demo codes can be found in the Github repository (https://github.com/NTU-Ling-lab/ORG-Classification)
QualGames: A Qualtrics implementation and a database of behavioral game theory tasks
No full textHere we share a dataset of responses from 314 participants in Singapore in 12 online decision-making tasks, and 94 participants in the United States in 10 decision-making tasks.
Demographics of the Singapore participants:
age: range [18.04, 29.98], mean 22.56, SD 2.16;
gender: 52.23% female;
ethnicity: 87.90% Chinese, 4.46% Malay, 5.10% Indian, and 2.55% other.
Individual-level demographics are not included to protect participants’ privacy.
The tasks administered to the Singapore participants are listed below:
Risk preference task (positive domain; PRp),
Risk preference task (negative domain; RPn),
Risk preference task (mixed domain; RPm),
Ambiguity aversion task (AA),
Social value orientation task (SVO),
Risky dictator game (RD),
Trust game (with no choice history; TGnh),
Trust game (with choice history; TG),
Trust game (as Player B; TGb),
Prisoner’s dilemma game (PD),
Stag hunt game (SH), and
Battle of the sexes game (BS).
Demographics of the U.S. participants:
age: range [21, 40], mean 32.48, SD 5.51;
gender: 35.11% female;
ethnicity: 58.51% White/Caucasian, 13.83% Hispanic, 14.89% Black/African American, 11.70% Asian/Pacific islander, and 1.06% American Indian/Alasca native.
The tasks administered to the U.S. participants are listed below:
Risk preference task (positive domain; PRp),
Risk preference task (negative domain; RPn),
Social value orientation task (SVO),
Risky dictator game (RD),
Trust game (with no choice history; TGnh),
Trust game (with choice history; TG),
Prisoner’s dilemma game (PD),
Ultimatum game (as proposer; UGp),
Ultimatum game (as responder; UGr), and
Temporal discounting task (TD).
These tasks were implemented on the Qualtrics platform. The details such as each screen that a participant may see when performing these tasks and the task code can be found in this GitHub repository:
https://github.com/ntu-cam-clic/Social_Decision_Making_Tasks.git</a