Vocapia Research

About: Vocapia Research is a company organization based out in Orsay, France. It is known for research contribution in the topics: Speaker recognition & Language model. The organization has 17 authors who have published 37 publications receiving 542 citations. The organization is also known as: Vecsys Research.

Score normalization and system combination for improved keyword spotting

Abstract: We present two techniques that are shown to yield improved Keyword Spotting (KWS) performance when using the ATWV/MTWV performance measures: (i) score normalization, where the scores of different keywords become commensurate with each other and they more closely correspond to the probability of being correct than raw posteriors; and (ii) system combination, where the detections of multiple systems are merged together, and their scores are interpolated with weights which are optimized using MTWV as the maximization criterion. Both score normalization and system combination approaches show that significant gains in ATWV/MTWV can be obtained, sometimes on the order of 8-10 points (absolute), in five different languages. A variant of these methods resulted in the highest performance for the official surprise language evaluation for the IARPA-funded Babel project in April 2013.

Unsupervised Speaker Identification using Overlaid Texts in TV Broadcast

Abstract: We propose an approach for unsupervised speaker identification in TV broadcast videos, by combining acoustic speaker diarization with person names obtained via video OCR from overlaid texts. Three methods for the propagation of the overlaid names to the speech turns are compared, taking into account the co-occurence duration between the speaker clusters and the names provided by the video OCR and using a task-adapted variant of the TF-IDF information retrieval coefficient. These methods were tested on the REPERE dry-run evaluation corpus, containing 3 hours of annotated videos. Our best unsupervised system reaches a F-measure of 70.2% when considering all the speakers, and 81.7% if anchor speakers are left out. By comparison, a mono-modal, supervised speaker identification system with 535 speaker models trained on matching development data and additional TV and radio data only provided a 57.5% F-measure when considering all the speakers and 45.7% without anchor.

Comparing Multi-Stage Approaches for Cross-Show Speaker Diarization

Abstract: Acoustic speaker diarization is investigated for situations where a collection of shows from the same source needs to be processed. In this case, the same speaker should receive the same label across all shows. We compare different architectures for cross-show speaker diarization: the obvious concatenation of all shows, a hybrid system combining first a local clustering stage followed by a global clustering stage, and an incremental system which processes the shows in a predefined order and updates the speaker models accordingly. This latter system being best suited to real applicative situations. These three strategies were compared to a baseline single-show system on a set of 46 ten-minutes samples of British English scientific podcasts.

KRISTINA: a knowledge-based virtual conversation agent

Abstract: We present an intelligent embodied conversation agent with linguistic, social and emotional competence Unlike the vast majority of the state-of-the-art conversation agents, the proposed agent is constructed around an ontology-based knowledge model that allows for flexible reasoning-driven dialogue planning, instead of using predefined dialogue scripts It is further complemented by multimodal communication analysis and generation modules and a search engine for the retrieval of multimedia background content from the web needed for conducting a conversation on a given topic The evaluation of the 1st prototype of the agent shows a high degree of acceptance of the agent by the users with respect to its trustworthiness, naturalness, etc The individual technologies are being further improved in the 2nd prototype

A Divide-and-Conquer Approach for Language Identification Based on Recurrent Neural Networks.

Abstract: This paper describes the design of an acoustic language recognition system based on BLSTM that can discriminate closely related languages and dialects of the same language. We introduce a Divide-and-Conquer (D&C) method to quickly and successfully train an RNN-based multi-language classifier. Experiments compare this approach to the straightforward training of the same RNN, as well as to two widely used LID techniques: a phonotactic system using DNN acoustic models and an i-vector system. Results are reported on two different data sets: the 14 languages of NIST LRE07 and the 20 closely related languages and dialects of NIST OpenLRE15. In addition to reporting the NIST Cavg metric which served as the primary metric for the LRE07 and OpenLRE15 evaluations, the EER and LER are provided. When used with BLSTM, the D&C training scheme significantly outperformed the classical training method for multi-class RNNs. On the OpenLRE15 data set, this method also outperforms classical LID techniques and combines very well with a phonotactic system.