Google has announced a neural machine translation (NMT) system that it says will reduce translation errors across its Google Translate service by between 55 percent and 85 percent, calling the achievement by its team of scientists a "significant milestone". Announced on the 10-year anniversary of Google Translate, Google Brain Team research scientists Quoc Le and Mike Schuster said Google has used machine intelligence to improve its image and speech-recognition systems, but had previously found it "challenging" to improve machine translation until now. "Ten years ago, we announced the launch of Google Translate, together with the use of Phrase-Based Machine Translation as the key algorithm behind this service," they said. "Today, we announce the Google Neural Machine Translation system (GNMT), which utilizes state-of-the-art training techniques to achieve the largest improvements to date for machine translation quality." Unlike the currently used phrase-based machine translation (PBMT) system -- which translates words and phrases independently within a sentence, and is notorious for its mistranslations -- neural machine translation considers the entire sentence as one unit to be translated.
We introduce and describe the results of a novel shared task on bandit learning for machine translation. The task was organized jointly by Amazon and Heidelberg University for the first time at the Second Conference on Machine Translation (WMT 2017). The goal of the task is to encourage research on learning machine translation from weak user feedback instead of human references or post-edits. On each of a sequence of rounds, a machine translation system is required to propose a translation for an input, and receives a real-valued estimate of the quality of the proposed translation for learning. This paper describes the shared task's learning and evaluation setup, using services hosted on Amazon Web Services (AWS), the data and evaluation metrics, and the results of various machine translation architectures and learning protocols.
Google's latest take on machine translation could make it easier for people to communicate with those speaking a different language, by translating speech directly into text in a language they understand. The team trained its system on hundreds of hours of Spanish audio with corresponding English text. After a learning period, Google's system produced a better-quality English translation of Spanish speech than one that transcribed the speech into written Spanish first. And text translation service Google Translate already uses neural networks on its most popular language pairs, which lets it analyse entire sentences at once to figure out the best written translation.
This paper reports experimental results of a high performance (real-time) memory-based translation. Memorybased translation is a new approach to machine translation which uses examples, or cases, of past translations to carry out translation of sentences. This idea is counter to traditional machine translation systems which rely on extensive use of rules in parsing, transfer and generation. Although, there are some preliminary reports on the superiority of the memory-based translation in terms of its scalability, quality of translation, and easiness of grammar writing, we have not seen any reports on its performance. This is perhaps, the first report discussing the feasibility and problems of the approach based on actual massively parallel implementation using real data. We also claim that the architecture of the IXM2 associative processor is highly suitable for memory-based translation tasks. Parsing performance of the memorybased translation system attained a few milliseconds per sentence.
OpenNMT is a industrial-strength, open-source (MIT) neural machine translation system utilizing the Torch mathematical toolkit. OpenNMT is used as provided in production by major translation providers. The system is designed to be simple to use and easy to extend, while maintaining efficiency and state-of-the-art translation accuracy. Simple general-purpose interface, requires only source/target files. Speed and memory optimizations for high-performance GPU training.