-
SEACrowd: A Multilingual Multimodal Data Hub and Benchmark Suite for Southeast Asian Languages
Authors:
Holy Lovenia,
Rahmad Mahendra,
Salsabil Maulana Akbar,
Lester James V. Miranda,
Jennifer Santoso,
Elyanah Aco,
Akhdan Fadhilah,
Jonibek Mansurov,
Joseph Marvin Imperial,
Onno P. Kampman,
Joel Ruben Antony Moniz,
Muhammad Ravi Shulthan Habibi,
Frederikus Hudi,
Railey Montalan,
Ryan Ignatius,
Joanito Agili Lopo,
William Nixon,
Börje F. Karlsson,
James Jaya,
Ryandito Diandaru,
Yuze Gao,
Patrick Amadeus,
Bin Wang,
Jan Christian Blaise Cruz,
Chenxi Whitehouse
, et al. (36 additional authors not shown)
Abstract:
Southeast Asia (SEA) is a region rich in linguistic diversity and cultural variety, with over 1,300 indigenous languages and a population of 671 million people. However, prevailing AI models suffer from a significant lack of representation of texts, images, and audio datasets from SEA, compromising the quality of AI models for SEA languages. Evaluating models for SEA languages is challenging due t…
▽ More
Southeast Asia (SEA) is a region rich in linguistic diversity and cultural variety, with over 1,300 indigenous languages and a population of 671 million people. However, prevailing AI models suffer from a significant lack of representation of texts, images, and audio datasets from SEA, compromising the quality of AI models for SEA languages. Evaluating models for SEA languages is challenging due to the scarcity of high-quality datasets, compounded by the dominance of English training data, raising concerns about potential cultural misrepresentation. To address these challenges, we introduce SEACrowd, a collaborative initiative that consolidates a comprehensive resource hub that fills the resource gap by providing standardized corpora in nearly 1,000 SEA languages across three modalities. Through our SEACrowd benchmarks, we assess the quality of AI models on 36 indigenous languages across 13 tasks, offering valuable insights into the current AI landscape in SEA. Furthermore, we propose strategies to facilitate greater AI advancements, maximizing potential utility and resource equity for the future of AI in SEA.
△ Less
Submitted 8 July, 2024; v1 submitted 14 June, 2024;
originally announced June 2024.
-
Ef-QuantFace: Streamlined Face Recognition with Small Data and Low-Bit Precision
Authors:
William Gazali,
Jocelyn Michelle Kho,
Joshua Santoso,
Williem
Abstract:
In recent years, model quantization for face recognition has gained prominence. Traditionally, compressing models involved vast datasets like the 5.8 million-image MS1M dataset as well as extensive training times, raising the question of whether such data enormity is essential. This paper addresses this by introducing an efficiency-driven approach, fine-tuning the model with just up to 14,000 imag…
▽ More
In recent years, model quantization for face recognition has gained prominence. Traditionally, compressing models involved vast datasets like the 5.8 million-image MS1M dataset as well as extensive training times, raising the question of whether such data enormity is essential. This paper addresses this by introducing an efficiency-driven approach, fine-tuning the model with just up to 14,000 images, 440 times smaller than MS1M. We demonstrate that effective quantization is achievable with a smaller dataset, presenting a new paradigm. Moreover, we incorporate an evaluation-based metric loss and achieve an outstanding 96.15% accuracy on the IJB-C dataset, establishing a new state-of-the-art compressed model training for face recognition. The subsequent analysis delves into potential applications, emphasizing the transformative power of this approach. This paper advances model quantization by highlighting the efficiency and optimal results with small data and training time.
△ Less
Submitted 28 February, 2024;
originally announced February 2024.
-
On Manipulating Scene Text in the Wild with Diffusion Models
Authors:
Joshua Santoso,
Christian Simon,
Williem
Abstract:
Diffusion models have gained attention for image editing yielding impressive results in text-to-image tasks. On the downside, one might notice that generated images of stable diffusion models suffer from deteriorated details. This pitfall impacts image editing tasks that require information preservation e.g., scene text editing. As a desired result, the model must show the capability to replace th…
▽ More
Diffusion models have gained attention for image editing yielding impressive results in text-to-image tasks. On the downside, one might notice that generated images of stable diffusion models suffer from deteriorated details. This pitfall impacts image editing tasks that require information preservation e.g., scene text editing. As a desired result, the model must show the capability to replace the text on the source image to the target text while preserving the details e.g., color, font size, and background. To leverage the potential of diffusion models, in this work, we introduce Diffusion-BasEd Scene Text manipulation Network so-called DBEST. Specifically, we design two adaptation strategies, namely one-shot style adaptation and text-recognition guidance. In experiments, we thoroughly assess and compare our proposed method against state-of-the-arts on various scene text datasets, then provide extensive ablation studies for each granularity to analyze our performance gain. Also, we demonstrate the effectiveness of our proposed method to synthesize scene text indicated by competitive Optical Character Recognition (OCR) accuracy. Our method achieves 94.15% and 98.12% on COCO-text and ICDAR2013 datasets for character-level evaluation.
△ Less
Submitted 3 November, 2023; v1 submitted 1 November, 2023;
originally announced November 2023.
-
Developing and implementing an Einsteinian science curriculum from Years 3 to 10: Part B Teacher upskilling: response to training and teacher's classroom experience
Authors:
Tejinder Kaur,
Magdalena Kersting,
Kyla Adams,
David Blair,
David Treagust,
Anastasia Popkova,
Shon Boublil,
Jesse Santoso,
Li Ju,
Marjan Zadnik,
David Wood,
Elaine Horne,
Darren McGoran,
Susan Scott,
Grady Venville
Abstract:
Recent years have seen a growing interest in modernizing physics and science curricula around the world. While many science educators and curriculum developers design instructional resources to successfully introduce topics of Einsteinian physics to young learners, it is clear that successful curriculum development needs to rest on successful teacher professional development.Teachers with or witho…
▽ More
Recent years have seen a growing interest in modernizing physics and science curricula around the world. While many science educators and curriculum developers design instructional resources to successfully introduce topics of Einsteinian physics to young learners, it is clear that successful curriculum development needs to rest on successful teacher professional development.Teachers with or without science backgrounds were trained in short professional learning workshops or completed micro-credential courses. The courses enabled teachers to gain knowledge and confidence to deliver the Einstein-First program. Detailed lesson plans and instructional videos for teachers define the lessons. Questionnaires were used to collect data, and teacher interviews were conducted following the various teacher training programs. The research results show that teachers effectively deliver the Einsteinian physics programs and that their subject matter and pedagogical content knowledge increased. In addition, teacher attitudes were favorable towards modernizing the physics curriculum. We conclude that it is feasible to upskill teachers from diverse backgrounds in Einsteinian physics and break the cycle that has inhibited the modernization of school curricula.
△ Less
Submitted 19 December, 2023; v1 submitted 29 June, 2023;
originally announced June 2023.
-
Developing and implementing an Einsteinian science curriculum from Years 3 to 10 : Part A Concepts, rationale and learning outcomes
Authors:
Tejinder Kaur,
Magdalena Kersting,
David Blair,
Kyla Adams,
David Treagust,
Jesse Santoso,
Anastasia Popkova,
Shon Boublil,
Marjan Zadnik,
Li Ju,
David Wood,
Elaine Horne,
Darren McGoran
Abstract:
There has been a growing realisation that school science curricula do not adequately reflect the revolutionary changes in our scientific understanding of the 20th century. This discrepancy between current school education and our modern scientific understanding has led to calls for the modernisation of the science curriculum. Although there have been attempts to introduce topics of Einsteinian phy…
▽ More
There has been a growing realisation that school science curricula do not adequately reflect the revolutionary changes in our scientific understanding of the 20th century. This discrepancy between current school education and our modern scientific understanding has led to calls for the modernisation of the science curriculum. Although there have been attempts to introduce topics of Einsteinian physics (i.e., quantum physics and relativity) to school education, often at the secondary level, we still lack a seamless curriculum in which modern science concepts are gradually introduced in primary and middle schools. Guided by the Model of Educational Reconstruction and following a mixed-methods research design, the Einstein-First project aims to address this gap. Einstein-First has developed and implemented an Einsteinian curriculum from Years 3 to 10 (students aged 7- 16) that resolves the disconnect between science in schools and the modern world. This paper presents the concepts, rationale, and learning outcomes of the curriculum implementation in six Australian schools with 315 students across Years 3 to 10. Our findings lay the foundation for informed curriculum development towards a school education that can enhance students' understanding and appreciation of the fundamental concepts of modern science and its impact on our society.
△ Less
Submitted 21 November, 2023; v1 submitted 29 June, 2023;
originally announced June 2023.
-
NusaCrowd: Open Source Initiative for Indonesian NLP Resources
Authors:
Samuel Cahyawijaya,
Holy Lovenia,
Alham Fikri Aji,
Genta Indra Winata,
Bryan Wilie,
Rahmad Mahendra,
Christian Wibisono,
Ade Romadhony,
Karissa Vincentio,
Fajri Koto,
Jennifer Santoso,
David Moeljadi,
Cahya Wirawan,
Frederikus Hudi,
Ivan Halim Parmonangan,
Ika Alfina,
Muhammad Satrio Wicaksono,
Ilham Firdausi Putra,
Samsul Rahmadani,
Yulianti Oenang,
Ali Akbar Septiandri,
James Jaya,
Kaustubh D. Dhole,
Arie Ardiyanti Suryani,
Rifki Afina Putri
, et al. (22 additional authors not shown)
Abstract:
We present NusaCrowd, a collaborative initiative to collect and unify existing resources for Indonesian languages, including opening access to previously non-public resources. Through this initiative, we have brought together 137 datasets and 118 standardized data loaders. The quality of the datasets has been assessed manually and automatically, and their value is demonstrated through multiple exp…
▽ More
We present NusaCrowd, a collaborative initiative to collect and unify existing resources for Indonesian languages, including opening access to previously non-public resources. Through this initiative, we have brought together 137 datasets and 118 standardized data loaders. The quality of the datasets has been assessed manually and automatically, and their value is demonstrated through multiple experiments. NusaCrowd's data collection enables the creation of the first zero-shot benchmarks for natural language understanding and generation in Indonesian and the local languages of Indonesia. Furthermore, NusaCrowd brings the creation of the first multilingual automatic speech recognition benchmark in Indonesian and the local languages of Indonesia. Our work strives to advance natural language processing (NLP) research for languages that are under-represented despite being widely spoken.
△ Less
Submitted 21 July, 2023; v1 submitted 19 December, 2022;
originally announced December 2022.
-
Market Potential for CO$_2$ Removal and Sequestration from Renewable Natural Gas Production in California
Authors:
Jun Wong,
Jonathan Santoso,
Marjorie Went,
Daniel Sanchez
Abstract:
Bioenergy with Carbon Capture and Sequestration (BECCS) is critical for stringent climate change mitigation, but is commercially and technologically immature and resource-intensive. In California, state and federal fuel and climate policies can drive first-markets for BECCS. We develop a spatially explicit optimization model to assess niche markets for renewable natural gas (RNG) production with c…
▽ More
Bioenergy with Carbon Capture and Sequestration (BECCS) is critical for stringent climate change mitigation, but is commercially and technologically immature and resource-intensive. In California, state and federal fuel and climate policies can drive first-markets for BECCS. We develop a spatially explicit optimization model to assess niche markets for renewable natural gas (RNG) production with carbon capture and sequestration (CCS) from waste biomass in California. Existing biomass residues produce biogas and RNG and enable low-cost CCS through the upgrading process and CO$_2$ truck transport. Under current state and federal policy incentives, we could capture and sequester 2.9 million MT CO$_2$/year (0.7% of California's 2018 CO$_2$ emissions) and produce 93 PJ RNG/year (4% of California's 2018 natural gas demand) with a profit maximizing objective. Existing federal and state policies produce profits of \$11/GJ. Distributed RNG production with CCS potentially catalyzes markets and technologies for CO$_2$ capture, transport, and storage in California.
△ Less
Submitted 4 May, 2021;
originally announced May 2021.
-
Measuring the jet power of flat spectrum radio quasars
Authors:
Stanislav Shabala,
Jesse Santoso,
Leith Godfrey
Abstract:
We use frequency-dependent position shifts of flat spectrum radio cores to estimate the kinetic power of AGN jets. We find a correlation between the derived jet powers and AGN narrow-line luminosity, consistent with the well-known relation for radio galaxies and steep spectrum quasars. This technique can be applied to intrinsically weak jets even at high redshift.
We use frequency-dependent position shifts of flat spectrum radio cores to estimate the kinetic power of AGN jets. We find a correlation between the derived jet powers and AGN narrow-line luminosity, consistent with the well-known relation for radio galaxies and steep spectrum quasars. This technique can be applied to intrinsically weak jets even at high redshift.
△ Less
Submitted 8 August, 2012;
originally announced August 2012.
-
Convexity of Momentum Maps: A Topological Analysis
Authors:
Wolfgang Rump,
Jenny Santoso
Abstract:
The Local-to-Global-Principle used in the proof of convexity theorems for momentum maps has been extracted as a statement of pure topology enriched with a structure of convexity. We extend this principle to not necessarily closed maps $f\colon X\ra Y$ where the convexity structure of the target space $Y$ need not be based on a metric. Using a new factorization of $f$, convexity of the image is pro…
▽ More
The Local-to-Global-Principle used in the proof of convexity theorems for momentum maps has been extracted as a statement of pure topology enriched with a structure of convexity. We extend this principle to not necessarily closed maps $f\colon X\ra Y$ where the convexity structure of the target space $Y$ need not be based on a metric. Using a new factorization of $f$, convexity of the image is proved without local fiber connectedness, and for arbitrary connected spaces $X$.
△ Less
Submitted 12 February, 2012; v1 submitted 9 November, 2010;
originally announced November 2010.
-
A General Local-to-Global Principle for Convexity of Momentum Maps
Authors:
Wolfgang Rump,
Jenny Santoso
Abstract:
We extend the Local-to-Global-Principle used in the proof of convexity theorems for momentum maps to not necessarily closed maps whose target space carries a convexity structure which need not be based on a metric. Using a new factorization of the momentum map, convexity of its image is proved without local fiber connectedness, and for almost arbitrary spaces of definition. Geodesics are obtained…
▽ More
We extend the Local-to-Global-Principle used in the proof of convexity theorems for momentum maps to not necessarily closed maps whose target space carries a convexity structure which need not be based on a metric. Using a new factorization of the momentum map, convexity of its image is proved without local fiber connectedness, and for almost arbitrary spaces of definition. Geodesics are obtained by straightening rather than shortening of arcs, which allows a unified treatment and extension of previous convexity results.
△ Less
Submitted 9 November, 2010; v1 submitted 13 September, 2010;
originally announced September 2010.
