Challenges in using a Standard Speech Recognition Engine in Small Vocabulary Domain

This paper discusses the challenges and proposes recommendations on using a standard speech recognition engine for a small vocabulary Air Traffic Controller Pilot communication domain. With the given challenges in transcribing the Air Traffic Communication due to the inherent radio issues in cockpit and the con-troller room, gathering the corpus for training the speech recognition model is another important problem. Taking advantage of the maturity of today’s speech recognition systems for the standard English words used in the communication, this paper focusses on the challenges in decoding the domain specific named entity words used in the communication.

Download Full-text

Domain-specific Evaluation Dataset Generator for Multilingual Text Analysis

Journal of Intelligent Systems with Applications ◽

10.54856/jiswa.201912084 ◽

2019 ◽

pp. 140-147

Author(s):

Emrah Inan ◽

Vahab Mostafapour ◽

Fatif Tekbacak

Keyword(s):

Text Analysis ◽

General Purpose ◽

Entity Linking ◽

Named Entity ◽

Domain Specific ◽

Benchmark Datasets ◽

Concise Information ◽

Multilingual Text ◽

The Given ◽

Specific Evaluation

Web enables to retrieve concise information about specific entities including people, organizations, movies and their features. Additionally, large amount of Web resources generally lies on a unstructured form and it tackles to find critical information for specific entities. Text analysis approaches such as Named Entity Recognizer and Entity Linking aim to identify entities and link them to relevant entities in the given knowledge base. To evaluate these approaches, there are a vast amount of general purpose benchmark datasets. However, it is difficult to evaluate domain-specific approaches due to lack of evaluation datasets for specific domains. This study presents WeDGeM that is a multilingual evaluation set generator for specific domains exploiting Wikipedia category pages and DBpedia hierarchy. Also, Wikipedia disambiguation pages are used to adjust the ambiguity level of the generated texts. Based on this generated test data, a use case for well-known Entity Linking systems supporting Turkish texts are evaluated in the movie domain.

Download Full-text

Machine Learning of Air Traffic Controller Command Extraction Models for Speech Recognition Applications

2020 AIAA/IEEE 39th Digital Avionics Systems Conference (DASC) ◽

10.1109/dasc50938.2020.9256484 ◽

2020 ◽

Author(s):

Hartmut Helmke ◽

Matthias Kleinert ◽

Oliver Ohneiser ◽

Heiko Ehr ◽

Shruthi Shetty

Keyword(s):

Machine Learning ◽

Speech Recognition ◽

Air Traffic ◽

Air Traffic Controller

Download Full-text

Automatic Speech Recognition for Air Traffic Control Communications

Transportation Research Record Journal of the Transportation Research Board ◽

10.1177/03611981211036359 ◽

2021 ◽

pp. 036119812110363

Author(s):

Sandeep Badrinath ◽

Hamsa Balakrishnan

Keyword(s):

Speech Recognition ◽

Language Processing ◽

Automatic Speech Recognition ◽

Traffic Control ◽

Air Traffic Control ◽

Air Traffic ◽

Machine Learning Techniques ◽

Air Traffic Controller ◽

Learning Techniques ◽

Operational Information

A significant fraction of communications between air traffic controllers and pilots is through speech, via radio channels. Automatic transcription of air traffic control (ATC) communications has the potential to improve system safety, operational performance, and conformance monitoring, and to enhance air traffic controller training. We present an automatic speech recognition model tailored to the ATC domain that can transcribe ATC voice to text. The transcribed text is used to extract operational information such as call-sign and runway number. The models are based on recent improvements in machine learning techniques for speech recognition and natural language processing. We evaluate the performance of the model on diverse datasets.

Download Full-text

Predictive Validity of a Selection Procedure for Air Traffic Controller Trainees

Aviation Psychology and Applied Human Factors ◽

10.1027/2192-0923/a000039 ◽

2013 ◽

Vol 3 (1) ◽

pp. 19-27 ◽

Cited By ~ 6

Author(s):

Yvonne Pecena ◽

Doris Keye ◽

Kristin Conzelmann ◽

Dietrich Grasshoff ◽

Peter Maschke ◽

...

Keyword(s):

Predictive Validity ◽

Selection Procedure ◽

Air Traffic ◽

Performance Criteria ◽

Training Phase ◽

Air Traffic Controller ◽

Training Performance ◽

Navigation Service ◽

German Aerospace ◽

Training Success

The job of an air traffic controller (ATCO) is very specific and demanding. The assessment of potential suitable candidates requires a customized and efficient selection procedure. The German Aerospace Center DLR conducts a highly selective, multiple-stage selection procedure for ab initio ATCO applicants for the German Air Navigation Service Provider DFS. Successful applicants start their training with a training phase at the DFS Academy and then continue with a unit training phase in live traffic. ATCO validity studies are scarcely reported in the international scientific literature and have mainly been conducted in a military context with only small and male samples. This validation study encompasses the data from 430 DFS ATCO trainees, starting with candidate selection and extending to the completion of their training. Validity analyses involved the prediction of training success and several training performance criteria derived from initial training. The final training success rate of about 79% was highly satisfactory and higher than that of other countries. The findings demonstrated that all stages of the selection procedure showed predictive validity toward training performance. Among the best predictors were scores measuring attention and multitasking ability, and ratings on general motivation from the interview.

Download Full-text