Research
Discover an index of datasets, SDKs, APIs and open-source tools developed by Microsoft researchers and shared with the global academic community below. These experimental technologies—available through Azure AI Foundry Labs (opens in new tab)—offer a glimpse into the future of AI innovation.
This code enables training and evaluation of a switching linear dynamic model for enhancing cepstral streams for automatic speech recognition, as described in our ICASSP 2004 paper, Noise Robust Speech Recognition with a Switching Linear…
Digital Effects, a plug-in for MSN Messenger video chatting, produces special effects for video communication. While the plug-in loads, users can select 3-D costumes or other digital effects in a pop-up window, and those effects…
The Orthant-Wise Limited-memory Quasi-Newton algorithm (OWL-QN) is a numerical optimization procedure for finding the optimum of an objective of the form {smooth function} plus {L1-norm of the parameters}. It has been used for training log-linear…
HTML and JavaScript Web code that displays a user interface to enable users to search, assimilate, and compare maps rapidly from an extensive corpus. The download includes a reference to a sample corpus of approximately…
Results from experiments conducted by Microsoft Research’s Machine Translation Incubation Team to investigate the impact of using good English (controlled language) on post-editing productivity—as well as on the overall quality of our statistical machine-translation system.
A set of manually annotated Chinese word-segmentation data and specifications for training and testing a Chinese word-segmentation system for research purposes. The data was extracted from the People’s Daily, which we have licensed for commercial…
This tool enables training and testing of maximum-entropy models using a general-feature file format. The tool also supports RProp and GIS as training algorithms.