33,466 research outputs found
State of the Art in Computational Bioacoustics and Machine Learning: How far have we come?
No full textTerrestrial bioacoustics, like many other domains, has recently witnessed some transformative results from the application of deep learning and big data (Stowell 2017, Mac Aodha et al. 2018, Fairbrass et al. 2018, Mercado III and Sturdy 2017). Generalising over specific projects, which bioacoustic tasks can we consider "solved"? What can we expect in the near future, and what remains hard to do? What does a bioacoustician need to understand about deep learning? This contribution will address these questions, giving the audience a concise summary of recent developments and ways forward. It builds on recent projects and evaluation campaigns led by the author (Stowell et al. 2015, Stowell et al. 2018), as well as broader developments in signal processing, machine learning and bioacoustic applications of these. We will discuss which type of deep learning networks are appropriate for audio data, how to address zoological/ecological applications which often have few available data, and issues in integrating deep learning predictions with existing workflows in statistical ecology
Datasets for automatic acoustic identification of individual birds
No full textBird individual audio recordings (foreground and background) to accompany the work:
"Automatic acoustic identification of individuals: Improving generalisation across species and recording conditions"
by Dan Stowell, Tereza Petrusková, Martin Šálek, Pavel Linhart
This dataset contains labelled recordings of individuals from three different bird species:
Little owl
Chiffchaff
Tree Pipit
For more information, please see the README.txt file, and the research article.
The dataset takes approx 11 GB of disk space after the ZIP files have been uncompressed.
</p
Automatic large-scale classification of bird sounds is strongly improved by unsupervised feature learning
Get PDFAutomatic species classification of birds from their sound is a computational tool of increasing importance in ecology, conservation monitoring and vocal communication studies. To make classification useful in practice, it is crucial to improve its accuracy while ensuring that it can run at big data scales. Many approaches use acoustic measures based on spectrogram-type data, such as the Mel-frequency cepstral coefficient (MFCC) features which represent a manually-designed summary of spectral information. However, recent work in machine learning has demonstrated that features learnt automatically from data can often outperform manually-designed feature transforms. Feature learning can be performed at large scale and “unsupervised”, meaning it requires no manual data labelling, yet it can improve performance on “supervised” tasks such as classification. In this work we introduce a technique for feature learning from large volumes of bird sound recordings, inspired by techniques that have proven useful in other domains. We experimentally compare twelve different feature representations derived from the Mel spectrum (of which six use this technique), using four large and diverse databases of bird vocalisations, classified using a random forest classifier. We demonstrate that in our classification tasks, MFCCs can often lead to worse performance than the raw Mel spectral data from which they are derived. Conversely, we demonstrate that unsupervised feature learning provides a substantial boost over MFCCs and Mel spectra without adding computational complexity after the model has been trained. The boost is particularly notable for single-label classification tasks at large scale. The spectro-temporal activations learned through our procedure resemble spectro-temporal receptive fields calculated from avian primary auditory forebrain. However, for one of our datasets, which contains substantial audio data but few annotations, increased performance is not discernible. We study the interaction between dataset characteristics and choice of feature representation through further empirical analysis
Making music through real-time voice timbre analysis: machine learning and timbral control
No full textPhDPeople can achieve rich musical expression through vocal sound { see for example
human beatboxing, which achieves a wide timbral variety through a range of
extended techniques. Yet the vocal modality is under-exploited as a controller
for music systems. If we can analyse a vocal performance suitably in real time,
then this information could be used to create voice-based interfaces with the
potential for intuitive and ful lling levels of expressive control.
Conversely, many modern techniques for music synthesis do not imply any
particular interface. Should a given parameter be controlled via a MIDI keyboard,
or a slider/fader, or a rotary dial? Automatic vocal analysis could provide
a fruitful basis for expressive interfaces to such electronic musical instruments.
The principal questions in applying vocal-based control are how to extract
musically meaningful information from the voice signal in real time, and how
to convert that information suitably into control data. In this thesis we address
these questions, with a focus on timbral control, and in particular we
develop approaches that can be used with a wide variety of musical instruments
by applying machine learning techniques to automatically derive the mappings
between expressive audio input and control output. The vocal audio signal is
construed to include a broad range of expression, in particular encompassing
the extended techniques used in human beatboxing.
The central contribution of this work is the application of supervised and
unsupervised machine learning techniques to automatically map vocal timbre
to synthesiser timbre and controls. Component contributions include a delayed
decision-making strategy for low-latency sound classi cation, a regression-tree
method to learn associations between regions of two unlabelled datasets, a fast
estimator of multidimensional di erential entropy and a qualitative method for
evaluating musical interfaces based on discourse analysis
Alaskan Author and Historian Dan O'Neill
No full textDan O'Neill has become a living legend in Alaska. He is the author of The Firecracker Boys: H-Bombs, Inupiat Eskimos, and the Roots of the Environmental Movement; A Land Gone Lonesome: An Inland Voyage Along the Yukon River; The Last Giant of Beringia: The Mystery of the Bering Land Bridge, and recently Stubborn Gal: The True Story of an Undefeated Sled Dog Racer, a children's book published by the University of Alaska Press. Dan came to Alaska in 1975 and has done a variety of things including dog mushing, trapping, hunting, working in construction, and on the pipeline. As research associate at the UAF Oral History program, he produced radio and television documentaries for public broadcasting, and for several years he wrote a column of political opinion for the Fairbanks Daily News-Miner
Bird Audio Detection Challenge 2016: public data
No full textData labels for the Bird Audio Challenge 2016. These annotations indicate the presence/absence of bird sounds in various datasets of 10-second audio clips.More info:http://machine-listening.eecs.qmul.ac.uk/bird-audio-detection-challenge/===============REUSE:The data is published under a Creative Commons CC-BY licence, and so you may reuse it as long as you attribute the origin. In academic work, please do this via a citation to our paper on the challenge:Dan Stowell, Mike Wood, Yannis Stylianou, Hervé Glotin. Bird detection in audio: a survey and a challenge, in Proceedings of MLSP 2016, arXiv:1608.03417 [cs.SD], 2016.You may also wish to link directly to this dataset, which you can do via its DOI https://dx.doi.org/10.6084/m9.figshare.3851466===============How the annotations were collected:The label "hasbird" represents whether a sound clip contains any audible bird sound. Note that there may be other sounds such as weather or humans, occasionally loud - the label does not mean that bird sound predominates.The labels were double-annotated, by the following procedures:* For the ff1010bird data, the presence/absence of birds was first deduced from the tags provided by the original FreeSound recordist (diverse crowdsourced tags such as "birdsong" "dawn-chorus" or "phylloscopus-collybita"), and then double-checked and refined by a manual annotator.* For the warblrb10k data, the annotations were crowdsourced via a web
service, answering the following question: "Can you hear any birds in
this clip?" Annotations were then double-checked by a manual annotator
to correct false-positives and -negatives.To download the corresponding audio data please see http://machine-listening.eecs.qmul.ac.uk/bird-audio-detection-challenge/<br
PERANCANGAN APLIKASI ELECTRONIC MEDICAL RECORD (EMR) PADA INSTALASI RAWAT INAP BERBASIS WEB
Get PDFPelayanan medik dewasa ini membutuhkan sistem yang lebih efektif dan efisien, baik dalam
penggunaan waktu, tenaga maupun sarana. Dalam pengelolaan rekam medik, kenyataan masih
umumnya penggunaan rekam medik manual yang dinilai tak lagi andal menangani data medik
melahirkan ide konversi rekam medik manual kertas ke rekam medik elektronik karena efektivitas dan
efisiensinya.
Penelitian ini bertujuan menciptakan aplikasi rekam medik elektronik yang lebih dikenal
sebagai EMR (Electronic Medical Record) dari rekam medik kertas di Instalasi Rawat Inap Rumah
Sakit Umum Ananda Salatiga. Rekam medik elektronik dirancang dengan membuat form-form isian
catatan-catatan medik dalam proses perawatan pasien selama dirawat. Data-data medik ini kemudian
disimpan dalam basis data sistem dan dikelola secara digital. Setiap kali pengisian data medik pada
form-form tertentu, sistem akan menghasilkan kode yang membawa informasi khusus.
Pada akhirnya, sistem akan menghasilkan deret kode ICD (International Statistical
Classification of Diseases and Related Health Problems) dari kode-kode yang dihasilkan pada
pengisian form-form catatan medik. Deretan kode-kode ini mampu menggambarkan perkembangan
kondisi pasien dan penanganan medik yang diberikan selama perawatan. Data-data medik yang
tersimpan dapat ditampilkan kembali dalam bentuk catatan medik digital.
Kata kunci: rekam medik, rawat inap, EMR, IC
IMPROVED MULTIPLE BIRDSONG TRACKING WITH DISTRIBUTION DERIVATIVE METHOD AND MARKOV RENEWAL PROCESS CLUSTERING
No full textDS & MP are supported by an EPSRC Leadership Fellowship EP/G007144/1
Correspondence regarding the possiblity of a Kephart Memorial
No full textThis 1968 correspondence, between Jackson E. Price and Dan Davis, discusses the possibility of “Memorial Center” to Horace Kephart (1862-1931), noted naturalist, woodsman, journalist, and author and promoter of the Great Smoky Mountains National Park
Analisis Tata Kelola Maritim Indonesia: Implementasi Visi Pemerintah Daerah Istimewa Yogyakarta
Get PDFIndonesia sebagai negara maritim memiliki banyak peluang dan ancaman. Hal tersebut memerlukan adanya perhatian yang lebih besar terhadap wilayah laut. Penelitian ini membahas strategi Pemerintah Indonesia untuk meningkatkan peluang ekonomi dan mengatasi tantangan melalui Poros Maritim Dunia. Tujuan riset ini adalah menganalisa kebijakan Pemerintah Indonesia dan Pemerintah Provinsi Daerah Istimewa Yogyakarta (DIY) dalam pengembangan wilayah pesisir menggunakan teori geopolitik dan teori kekuatan laut. Penelitian ini menerapkan metode kualitatif studi kasus. Data yang digunakan berupa data primer melalui wawancara dan data sekunder yang diperoleh dari buku, artikel jurnal dan publikasi daring. Penulis menemukan bahwa sinergitas penegak hukum, diplomasi maritim sudah cukup baik, namun perlu lebih aktif melibatkan masyarakat untuk pengelolaan sumber daya hayati yang berkelanjutan dan peningkatan SDM masyarakat pesisir. Sementara Visi Gubernur DIY bernama Abad Samudera Hindia belum terlaksana dengan maksimal. Masyarakat membutuhkan tambahan pembangunan TPI baru dan pengembangan TPI menjadi PPN dan PPS untuk meningkatkan produktivitas dan kesejahteraan nelayan. Dapat disimpulkan bahwa pemerintah pusat dan daerah harus selalu memperhatikan pengelolaan laut, khususnya untuk pemanfaatan sumber daya dan pangan. Bahwa laut memiliki kekayaan alam yang berlimpah dapat digunakan meningkatkan gizi masyarakat dan pendapatan negara. Title: Indonesian Maritime Governance Analysis: Implementation of the Vision of the Yogyakarta Special Region GovernmentIndonesia, as a maritime country, has many opportunities and threats. This condition requires greater attention to the sea area. This research discusses the strategy of the Government of Indonesia to increase economic opportunities and overcome challenges through the World Maritime Fulcrum. This research aims to analyze the policies of the Government of Indonesia and the Provincial Government of DIY in developing coastal areas using geopolitical theory and the theory of sea power. This study uses a qualitative case study method. The data used are primary data through interviews and secondary data obtained from books, journal articles and online publications. The author finds that the synergism between law enforcement and maritime diplomacy is good enough. However, it is necessary to involve the community more actively in the sustainable management of biological resources and to increase the human resources of coastal communities. Meanwhile, the vision of the Governor of DIY called the Century of the Indian Ocean, has not been implemented optimally. The community needs additional construction of new TPI and development of TPI to become PPN and PPS to increase the productivity and welfare of fishermen. It can be concluded that the central and regional governments must always pay attention to marine management, especially for utilising resources and food. That the sea has abundant natural wealth can be used to improve people’s nutrition and state income
- …
