You are here

HLT-mt events

Subscribe to HLT-mt events feed
Updated: 1 min 9 sec ago

HLT-MT @ Researchers' Night

The HLT-MT unit will be present at the Researcher's Night event. The general theme of our boot will machine translation, with a particular focus and the "word alignment technology" used to train statistical machine translation systems.  The even will be attended by Luisa, Matteo and Marcello.

Language English

CloudASR: A Web Platform for Automatic Speech Recognition

In this talk I will present a cloud platform for automatic speech recognition, CloudASR, built on top of Kaldi speech recognition toolkit. The platform supports both batch and online speech recognition mode and it has an annotation interface for transcription of the submitted recordings. The key features of the platform are scalability, customizability and easy deployment. Benchmarks of the platform show that the platform achieves comparable performance with Google Speech API in terms of latency and it can achieve better accuracy on limited domains.

Language English

Automatic detection of event factuality in Italian

Presentation of internship work on event factuality detection in Italian news

Language English

Speech Production in Automatic Speech Recognition Systems

This presentation will focus on ASR combininig acoustic data with information about the vocal tract movements during speech production

Language English

Seminar on the use of Kaldi

Recent results achieved on the TED talk transcription task with Kaldi

Language English

MT Marathon 2014

The Machine Translation Marathon 2014 is a week long event which will include MT lectures and labs, talks about open source tools for MT, and open source MT hacking projects.

Language English

A Computational System for Poetry Style Analysis and an Expressive Poetry Reader

We present a system that computes an extended number of linguistic parameters to evaluate a poem, a set or collection of poems and compares the results.

Language English

Enriching the Neural Network Language Model with Linguistic Information for Improving the ASR output

In this work, we introduce two solutions for two major issues: 1) How to, effectively, use the Neural Network Language Model (NNLM) in ASR decoder; 2) How to improve the NNLM performance using linguistic information. For the former, we propose rescoring the theories in the stack of A* search, instead of rescoring the final N-best lists; and for the latter we propose a new deep structure for representing the linguistic information.

Language English

Advances in Word Reordering Modeling for Phrase-Based Statistical Machine Translation

We discuss the word reordering issue in phrase-based SMT and present some approaches to enhance the standard word reordering constraints

Language English

Bridging NLP and semantic web to enhance user interactions with the web of data

We address the development of methods for a flexible mapping between natural language expressions, and concepts and relations in structured knowledge bases

Language English

Learning Corpus Patterns Using Finite State Automata

We argue that natural language has computationally discoverable regular properties characterizing a certain type of phrases

Language English

Context-based Language Model Adaptation for Lecture Translation

We focus on topic adaptation for language modeling to improve the fluency of translations, both through word choice and small reordering decisions

Language English