Paper Info
Title
Poseidon - Passive-Acoustic Ocean Sensor For Entertainment And Interactive Data-Gathering In Opportunistic Nautical-Activities
Abstract
Recent years demonstrate an increased interest in low-cost Passive Acoustic Monitoring (PAM) in citizen science for ecological monitoring and environmental protection. However, most efforts have targeted land use, leaving ocean and nautical applications greatly unexplored. In this paper we present the design, deployment and testing of POSEIDON, a low-cost PAM system for nautical citizen science and real-time acoustic augmentation of whale-watching experiences. POSEIDON uses machine learning techniques to identify vocal acoustic samples of common cetaceans like whales and dolphins. When discriminating the calls, we find that Extra Trees and Gradient Boosting outperform other classifiers (>0.95 confidence threshold). The features extracted from the machine learning models are used to enhance the whale watching experience and provide citizen science data to marine biologists and environmental protection agencies. While this paper focuses on the design of the system, future work will focus on user testing and widespread deployment of open-hardware and software for nautical PAM applications.
Year
DOI
Venue
2018
10.1145/3196709.3196752
DIS 2018: PROCEEDINGS OF THE 2018 DESIGNING INTERACTIVE SYSTEMS CONFERENCE
Keywords
Field
DocType
Passive Acoustic Monitoring, Ocean GUI, Citizen Science, Cetaceans, Machine Learning, Whale-watching, IoT
Data collection,User testing,Whale watching,Software deployment,Entertainment,Human–computer interaction,Citizen science,Engineering,Multimedia,Gradient boosting,Underwater
Conference
Citations 
PageRank 
References 
0
0.34
9
Authors
4
Name
Order
Citations
PageRank
Marko Radeta121.43
Nuno Jardim Nunes242574.01
Dinarte Vasconcelos300.34
Valentina Nisi416036.54