Indexing and Retrieving Voice Recordings by Instantly Tagging Mentioned Objects with Dots

Thibaut van Bergen,Rui Ishiyama,Kengo Makino,Yuta Kudo,Toru Takahashi,Hans Goosen

The Internet of Things (IOT)（2019）

Cited 2|Views31

Abstract

This paper presents a novel framework and its prototype tool for indexing and retrieving specific fragments of voice recordings obtained during discussions about physical objects such as text documents, pictures, or 3D models. When a specific part of an object is mentioned, it is tagged with an ink dot that is immediately registered in a database by capturing a microscopic image of the dot. Simultaneously, an index of the recording fragment is created and linked with the dot. After the recording, a dot can be scanned and identified by matching its microscopic image with the database to retrieve the linked recording fragment for playback. A handy tool was developed to facilitate these operations while the user concentrates on the ongoing discussion. Performance tests of the dot identification have shown genuine matches without error. In demonstrations of a realistic usage scenario, the tool successfully facilitated the creation of indexes with dots during a voice recording and correctly played back all the specific recording fragments linked to the dots.

Translated text

Key words

voice recording,indexing,retrieval,playback,pattern recognition,identification,image matching

Bibtex

AI Read Science

AI Summary

AI Summary is the key point extracted automatically understanding the full text of the paper, including the background, methods, results, conclusions, icons and other key content, so that you can get the outline of the paper at a glance.

Example

Background

Key content

Introduction

Methods

Results

Related work

Fund

Key content

Pretraining has recently greatly promoted the development of natural language processing (NLP)
We show that M6 outperforms the baselines in multimodal downstream tasks, and the large M6 with 10 parameters can reach a better performance
We propose a method called M6 that is able to process information of multiple modalities and perform both single-modal and cross-modal understanding and generation
The model is scaled to large model with 10 billion parameters with sophisticated deployment, and the 10 -parameter M6-large is the largest pretrained model in Chinese
Experimental results show that our proposed M6 outperforms the baseline in a number of downstream tasks concerning both single modality and multiple modalities We will continue the pretraining of extremely large models by increasing data to explore the limit of its performance

Try using models to generate summary,it takes about 60s

Must-Reading Tree

Example

Generate MRT to find the research sequence of this paper