Extensive 30 years of experience in science and technology corporate management, human resource development, technology assessment and application, engineering, and innovation ecosystem. Leadership and organizational know-how with an emphasis on research commercialization, by focusing on Artificial Intelligence, digital transformation, enterprise architecture, cyber security, and communication systems.
Dr. Riza has a Doctor of Engineering degree in Electrical Engineering (cum laude) from Bandung Institute of Technology (ITB) Supervisors: Principal Engineer and Advisor
Applied Computational Intelligence and Soft Computing
Currently, speech recognition datasets are increasingly available freely in various languages. Ho... more Currently, speech recognition datasets are increasingly available freely in various languages. However, speech recognition datasets in the Indonesian language are still challenging to obtain. Consequently, research focusing on speech recognition is challenging to carry out. This research creates Indonesian speech recognition datasets from YouTube channels with subtitles by validating all utterances of downloaded audio to improve the data quality. The quality of the dataset was evaluated using a deep neural network. The time delay neural network (TDNN) was used to build the acoustic model by applying the alignment data from the Gaussian mixture model-hidden Markov model (GMM-HMM). Data augmentation was used to increase the number of validated datasets and enhance the performance of the acoustic model. The results show that the acoustic model built using the validated datasets is better than the unvalidated datasets for all types of lexicons. Utilizing the four lexicon types and incre...
2019 22nd Conference of the Oriental COCOSDA International Committee for the Co-ordination and Standardisation of Speech Databases and Assessment Techniques (O-COCOSDA)
With competencies and the results of the engineering of natural language processing technology ow... more With competencies and the results of the engineering of natural language processing technology owned by BPPT since 1987, BPPT develops an English-Bahasa Indonesia speech-to-speech translation system (S2ST). In this paper, we propose an architecture of speech-to-speech translation system for Android-based mobile conversation using separate mobile devices for each language. This architecture applies three leading technologies, namely: WebSocket, REST, and JSON. The system utilizes a two-way communication protocol between two users and a simple voice activation detector that can detect a boundary of user's utterance.
2017 20th Conference of the Oriental Chapter of the International Coordinating Committee on Speech Databases and Speech I/O Systems and Assessment (O-COCOSDA), 2017
This paper describes our natural language resources especially text and speech corpora for develo... more This paper describes our natural language resources especially text and speech corpora for developing an Indonesian speech-to-speech translation (S2ST) system. The corpora are used to create models for Automatic Speech Recognition (ASR), Statistical Machine Translation (SMT), and Text-to-Speech (TTS) systems. The corpora collected since 1987 from various sources and projects such as Multilingual Machine Translation System (MMTS), PAN Localization, ASEAN MT, U-STAR, etc. Text corpora are created by either collecting from online resources or translating manually from textual sources. Speech corpora are made from several recording projects. Availability of these corpora enables us to develop Indonesian speech-to- speech translation system.
The emergence of a new variant of the coronavirus, SARS-Cov-2, which causes the Corona Virus Dise... more The emergence of a new variant of the coronavirus, SARS-Cov-2, which causes the Corona Virus Disease (Covid-19) outbreak has really changed the world. First reported in Wuhan City, Hubei Province, China at the end of 2019, this virus has spread throughout the world. Apart from hitting the world economy, the Covid-19 pandemic has also changed the way humans interact. All over the world, people have changed their habits of work, worship, and social activities. This was done to reduce the risk of transmission of the massive new coronavirus. But the next question arises: when will conditions improve? when will this Covid-19 outbreak subside? To answer this question, this study seeks to model the spread of the new Corona Virus with a Dynamic Systems approach. In the modelling carried out, there are seven scenarios that describe the policies undertaken to mitigate the spread of Covid-19 which include WFH policies, office vacations, social distancing, implementation of PSBB, to PSBB relaxa...
Statistical Machine Translation (SMT) model has limitations on mapping phrases or blocks of the s... more Statistical Machine Translation (SMT) model has limitations on mapping phrases or blocks of the source language to the target without the use of linguistic information. We can add part-of-speech (PoS) information as one of the linguistic features to improve the quality of translations. Indonesian PoS tagsets that are used to process natural language computing is very diverse, so we experimented to determine the best PoS tagset used as additional linguistic information on SMT. This paper discuss various PoS tag information as a feature in the SMT factored translation model, where we experiment using Moses and BLEU as an evaluation tool. We use several PoS tagset from computational linguistic studies in Indonesia. The experimental result shows that , Wicaksono's PoS tagset give a better BLEU score than the other PoS tagsets. This will enable the improvement of English-Indonesian SMT as part of our participation in the network-based ASEAN-MT system.
Speech-related research in Indonesia has focused mainly on speech synthesizer technologies and na... more Speech-related research in Indonesia has focused mainly on speech synthesizer technologies and natural language processing. More recently, work on collecting Indonesian speech corpora and developing speech recognition systems was initiated. However, most of these systems only recognized a very limited vocabulary. This paper outlines recent progress in developing Indonesian large-vocabulary corpora and a large vocabulary continuous speech recognition (LVCSR) system. Research on the Indonesian LVCSR has been carried out under the A-STAR (Asian Speech Translation Advanced Research) consortium. Three types of Indonesian large vocabulary data sources were used: daily news, telephone application and basic travel expression (BTEC) tasks, which are available in both text and speech forms. The Indonesian speech recognition engine was trained using clean speech for both the daily news and telephone application tasks, and the performance was evaluated using the BTEC task.
2009 IEEE Workshop on Automatic Speech Recognition & Understanding, 2009
Page 1. The Asian Network-based Speech-to-Speech Translation System Sakriani Sakti 1, Noriyuki Ki... more Page 1. The Asian Network-based Speech-to-Speech Translation System Sakriani Sakti 1, Noriyuki Kimura 1, Michael Paul 1, Chiori Hori 1, Eiichiro Sumita 1, Satoshi Nakamura 1, Jun Park2, Chai Wutiwiwatchai3, Bo Xu4, Hammam ...
2021 7th Asia-Pacific Conference on Synthetic Aperture Radar (APSAR), 2021
Forest fires, especially on peatlands, constitute a significant problem that is often faced in In... more Forest fires, especially on peatlands, constitute a significant problem that is often faced in Indonesia. Forest fires that occur on peatlands are caused because the peatlands have been degraded due to deforestation, with the aim of utilizing peatlands for forestry and agricultural cultivation activities. Efforts to mitigate forest fires on peatlands need to be made. The first thing to do is to map peatlands that have a high risk of burning. In order to cover a large area, remote sensing technology is needed. Peatlands that have subsided are identical to peatlands that are dry, so they burn easily. SAR interferometry has the ability to detect the subsidence of peatlands. Combined with the Wosten model, areas with a high potential risk of forest fire on peatland can be mapped, and this has been demonstrated in this research.
2020 International Conference on ICT for Smart Society (ICISS), 2020
The fourth industrial revolution essential components which include cloud computing technology, a... more The fourth industrial revolution essential components which include cloud computing technology, artificial intelligence, big data, and the internet of things has also been affecting the flood disaster mitigation strategy worldwide. This is also true for Indonesia where the flood disaster event has increased almost three times during the past fifteen years. In this literature review, the advancement of the application of artificial intelligence, in particular, machine learning in flood mitigation in Indonesia is studied. Based on this study, the future possible improvement of flood mitigation has also been given. The study revealed that machine learning has not yet been applied extensively in flood disaster mitigation in Indonesia. Some applications are for rainfall, river water level, and discharge level predictions, and to a lesser degree for flood forecasting and early warning. The future prospective advancement of flood disaster mitigation is the application of machine learning methods in an integrated flood prediction and early warning that covers rainfall, river water level, river discharge, and flood predictions as well as estimated spatial flooding area, and early warning dissemination using the internet of things.
In this paper, we present a report on the re-search and development of speech to speech translati... more In this paper, we present a report on the re-search and development of speech to speech translation system for Asian lan-guages, primarily on the design and im-plementation of speech recognition and machine translation systems for Indonesia language. As part of the A-STAR project, each participating country will need to de-velop each component of the full system for the corresponding language. We will specifically discuss our method on building speech recognition and stochastic language model for statistically translating Indone-sian into other Asian languages. The sys-tem is equipped with a capability to handle variation of speech input, a more natural mode of communication between the sys-tem and the users. 1
Applied Computational Intelligence and Soft Computing
Currently, speech recognition datasets are increasingly available freely in various languages. Ho... more Currently, speech recognition datasets are increasingly available freely in various languages. However, speech recognition datasets in the Indonesian language are still challenging to obtain. Consequently, research focusing on speech recognition is challenging to carry out. This research creates Indonesian speech recognition datasets from YouTube channels with subtitles by validating all utterances of downloaded audio to improve the data quality. The quality of the dataset was evaluated using a deep neural network. The time delay neural network (TDNN) was used to build the acoustic model by applying the alignment data from the Gaussian mixture model-hidden Markov model (GMM-HMM). Data augmentation was used to increase the number of validated datasets and enhance the performance of the acoustic model. The results show that the acoustic model built using the validated datasets is better than the unvalidated datasets for all types of lexicons. Utilizing the four lexicon types and incre...
2019 22nd Conference of the Oriental COCOSDA International Committee for the Co-ordination and Standardisation of Speech Databases and Assessment Techniques (O-COCOSDA)
With competencies and the results of the engineering of natural language processing technology ow... more With competencies and the results of the engineering of natural language processing technology owned by BPPT since 1987, BPPT develops an English-Bahasa Indonesia speech-to-speech translation system (S2ST). In this paper, we propose an architecture of speech-to-speech translation system for Android-based mobile conversation using separate mobile devices for each language. This architecture applies three leading technologies, namely: WebSocket, REST, and JSON. The system utilizes a two-way communication protocol between two users and a simple voice activation detector that can detect a boundary of user's utterance.
2017 20th Conference of the Oriental Chapter of the International Coordinating Committee on Speech Databases and Speech I/O Systems and Assessment (O-COCOSDA), 2017
This paper describes our natural language resources especially text and speech corpora for develo... more This paper describes our natural language resources especially text and speech corpora for developing an Indonesian speech-to-speech translation (S2ST) system. The corpora are used to create models for Automatic Speech Recognition (ASR), Statistical Machine Translation (SMT), and Text-to-Speech (TTS) systems. The corpora collected since 1987 from various sources and projects such as Multilingual Machine Translation System (MMTS), PAN Localization, ASEAN MT, U-STAR, etc. Text corpora are created by either collecting from online resources or translating manually from textual sources. Speech corpora are made from several recording projects. Availability of these corpora enables us to develop Indonesian speech-to- speech translation system.
The emergence of a new variant of the coronavirus, SARS-Cov-2, which causes the Corona Virus Dise... more The emergence of a new variant of the coronavirus, SARS-Cov-2, which causes the Corona Virus Disease (Covid-19) outbreak has really changed the world. First reported in Wuhan City, Hubei Province, China at the end of 2019, this virus has spread throughout the world. Apart from hitting the world economy, the Covid-19 pandemic has also changed the way humans interact. All over the world, people have changed their habits of work, worship, and social activities. This was done to reduce the risk of transmission of the massive new coronavirus. But the next question arises: when will conditions improve? when will this Covid-19 outbreak subside? To answer this question, this study seeks to model the spread of the new Corona Virus with a Dynamic Systems approach. In the modelling carried out, there are seven scenarios that describe the policies undertaken to mitigate the spread of Covid-19 which include WFH policies, office vacations, social distancing, implementation of PSBB, to PSBB relaxa...
Statistical Machine Translation (SMT) model has limitations on mapping phrases or blocks of the s... more Statistical Machine Translation (SMT) model has limitations on mapping phrases or blocks of the source language to the target without the use of linguistic information. We can add part-of-speech (PoS) information as one of the linguistic features to improve the quality of translations. Indonesian PoS tagsets that are used to process natural language computing is very diverse, so we experimented to determine the best PoS tagset used as additional linguistic information on SMT. This paper discuss various PoS tag information as a feature in the SMT factored translation model, where we experiment using Moses and BLEU as an evaluation tool. We use several PoS tagset from computational linguistic studies in Indonesia. The experimental result shows that , Wicaksono's PoS tagset give a better BLEU score than the other PoS tagsets. This will enable the improvement of English-Indonesian SMT as part of our participation in the network-based ASEAN-MT system.
Speech-related research in Indonesia has focused mainly on speech synthesizer technologies and na... more Speech-related research in Indonesia has focused mainly on speech synthesizer technologies and natural language processing. More recently, work on collecting Indonesian speech corpora and developing speech recognition systems was initiated. However, most of these systems only recognized a very limited vocabulary. This paper outlines recent progress in developing Indonesian large-vocabulary corpora and a large vocabulary continuous speech recognition (LVCSR) system. Research on the Indonesian LVCSR has been carried out under the A-STAR (Asian Speech Translation Advanced Research) consortium. Three types of Indonesian large vocabulary data sources were used: daily news, telephone application and basic travel expression (BTEC) tasks, which are available in both text and speech forms. The Indonesian speech recognition engine was trained using clean speech for both the daily news and telephone application tasks, and the performance was evaluated using the BTEC task.
2009 IEEE Workshop on Automatic Speech Recognition & Understanding, 2009
Page 1. The Asian Network-based Speech-to-Speech Translation System Sakriani Sakti 1, Noriyuki Ki... more Page 1. The Asian Network-based Speech-to-Speech Translation System Sakriani Sakti 1, Noriyuki Kimura 1, Michael Paul 1, Chiori Hori 1, Eiichiro Sumita 1, Satoshi Nakamura 1, Jun Park2, Chai Wutiwiwatchai3, Bo Xu4, Hammam ...
2021 7th Asia-Pacific Conference on Synthetic Aperture Radar (APSAR), 2021
Forest fires, especially on peatlands, constitute a significant problem that is often faced in In... more Forest fires, especially on peatlands, constitute a significant problem that is often faced in Indonesia. Forest fires that occur on peatlands are caused because the peatlands have been degraded due to deforestation, with the aim of utilizing peatlands for forestry and agricultural cultivation activities. Efforts to mitigate forest fires on peatlands need to be made. The first thing to do is to map peatlands that have a high risk of burning. In order to cover a large area, remote sensing technology is needed. Peatlands that have subsided are identical to peatlands that are dry, so they burn easily. SAR interferometry has the ability to detect the subsidence of peatlands. Combined with the Wosten model, areas with a high potential risk of forest fire on peatland can be mapped, and this has been demonstrated in this research.
2020 International Conference on ICT for Smart Society (ICISS), 2020
The fourth industrial revolution essential components which include cloud computing technology, a... more The fourth industrial revolution essential components which include cloud computing technology, artificial intelligence, big data, and the internet of things has also been affecting the flood disaster mitigation strategy worldwide. This is also true for Indonesia where the flood disaster event has increased almost three times during the past fifteen years. In this literature review, the advancement of the application of artificial intelligence, in particular, machine learning in flood mitigation in Indonesia is studied. Based on this study, the future possible improvement of flood mitigation has also been given. The study revealed that machine learning has not yet been applied extensively in flood disaster mitigation in Indonesia. Some applications are for rainfall, river water level, and discharge level predictions, and to a lesser degree for flood forecasting and early warning. The future prospective advancement of flood disaster mitigation is the application of machine learning methods in an integrated flood prediction and early warning that covers rainfall, river water level, river discharge, and flood predictions as well as estimated spatial flooding area, and early warning dissemination using the internet of things.
In this paper, we present a report on the re-search and development of speech to speech translati... more In this paper, we present a report on the re-search and development of speech to speech translation system for Asian lan-guages, primarily on the design and im-plementation of speech recognition and machine translation systems for Indonesia language. As part of the A-STAR project, each participating country will need to de-velop each component of the full system for the corresponding language. We will specifically discuss our method on building speech recognition and stochastic language model for statistically translating Indone-sian into other Asian languages. The sys-tem is equipped with a capability to handle variation of speech input, a more natural mode of communication between the sys-tem and the users. 1
Uploads