Articles tagged with Machine Translation
A recent study from UTSA researchers reveals that large language models (LLMs) can pose a serious threat to programmers who use them to help write code. The study found that up to 97% of software developers incorporate generative AI into their workflow, and 30% of code written today is AI-generated.
A new method has improved AI translation of sign language by adding data on hand and facial expressions, as well as skeletal information. This has led to a significant increase in accuracy, making it easier for people with hearing impairments to communicate.
A new paper in JCOM describes the development of an automatic translation app for sign language, designed through co-creation with deaf communities. The research team used a theatrical performance and AI tools to gather feedback from audiences, informing the app's design and features.
A novel, fast and high-quality neural text-to-speech model was successfully developed using a Transformer encoder + ConvNeXt decoder and MS-FC-HiFi-GAN. The model can synthesize one second of speech at high speed in just 0.1 seconds using a single CPU core, achieving eight times faster synthesis than conventional methods.
Anastasopoulos is helping to develop a solution that can automatically translate languages of the southeast Asia and Pacific regions, with a particular focus on Indonesian and Filipino languages. Funding for this project began in January 2024 and will end in July 2024.
The JGU Center for Lifelong Learning is developing personalized AI-based learning experiences for adult learners, aiming to improve motivation and learning outcomes. The project will also discuss the challenges of AI technologies in teaching modern foreign languages.
A new study published in Frontiers in Artificial Intelligence found that training in Interlingual Respeaking, a new practice combining human collaboration with speech recognition software, can improve language professionals' cognitive abilities. The research, conducted by the University of Surrey, showed significant enhancements in wor...
A new machine learning model can automatically translate Akkadian text written in cuneiform into English, with the first version using Latin transliteration achieving satisfactory results. The program is effective for translating short sentences and can be used as part of a human-machine collaboration to correct and refine its output.
A new study at Cornell University introduced an automated participant that periodically interrupts the conversation to give nonnative speakers a chance to speak. The AI bot increased participation from 12% to 17% of all words spoken, with nonnative speakers feeling valued and appreciated for their perspectives.
A research team at Carnegie Mellon University is working to simplify data requirements for speech recognition models, aiming to reach 2,000 languages. By focusing on linguistic elements common across many languages and using a phylogenetic tree, the team hopes to eliminate the need for audio data.
A new paper introduces free machine translation services to convert scientific papers into multiple languages, promoting accessibility and global collaboration. Researchers at UC Berkeley developed these tools to address the dominance of English in scientific research, enabling scientists worldwide to share their work.
A team of researchers from Skoltech and universities developed a neural network-based solution for automated recognition of chemical formulas on research paper scans. The algorithm combines molecules, functional groups, fonts, styles, and printing defects to mimic existing molecular template depiction styles.
The University of Tartu and Tilde are developing a cutting-edge machine translation platform to advance Estonian translation capabilities. The project aims to create practical tools for public institutions, translation agencies, and developers, leveraging new approaches like modular transformer-type neural networks.
Researchers are creating a machine translation program for web browsers like Mozilla Firefox, focusing on improving translation quality and ensuring user privacy. The project involves adapting the engine to context, recognizing nuances in language style, and estimating translation quality.
Researchers at NAIST created a deep learning-based system to transcribe Japanese lecture speech and translate it into English with near-realtime accuracy. The system uses archived lecture videos with subtitles in both languages, achieving better translations than traditional live translation methods.
Researchers found Google Translate to be 92% accurate for Spanish and 81% accurate for Chinese, with most errors due to grammar or typographical mistakes. Clinicians are advised to verify translations with human interpreters and provide patients with the original English instructions.
Researchers at Dartmouth College used the Bible to develop an algorithm that can convert written works into different styles for different audiences. The study, published in Royal Society Open Science, trained on over 1.5 million unique pairings of source and target verses from various versions of the sacred texts.
Researchers will use machine learning to analyze language patterns and morphology in low-resource languages like Tagalog and Swahili. The system aims to provide English summaries explaining how documents are relevant to the query.
Researchers used a newly developed interpretive technique to analyze neural networks trained for machine translation and speech recognition. They found that lower-level tasks, such as sound recognition or part-of-speech recognition, are prioritized before higher-level tasks like transcription or semantic interpretation.
Columbia University's SCRIPTS system uses AI to process documents in low-resource languages, providing summaries and translations. The project aims to enhance efficiency for intelligence analysts worldwide.
A new multilingual search tool developed at the University of Washington's Turing Center allows people to search for images on the Web using hundreds of languages. The tool, named PanImages, automatically translates the search term into about 300 other languages and displays images from Google and Flickr.
The new version boasts improved translation quality, speed improvements of 5-10 times over previous versions, and enhanced multi-processor capabilities. This upgrade enables faster automation of the translation process, saving customers time and money.
Researchers at USC's Information Sciences Institute built a state-of-the-art machine translation system for Hindi in less than a month, achieving all aspects of the DARPA's 'Surprise Language' project. The system uses statistical models to find the most likely translations and can query Hindi databases using English questions.