Visual Active Learning for Labeling: A Case for Soundscape Ecology Data

Autor(es)

Hilasaca, Liz Huancapaza ^[1] ; Ribeiro, Milton Cezar ^[2] ; Minghim, Rosane ^[3]

Afiliação do(s) autor(es)

^[1] Univ Sao Paulo, Dept Comp Sci, BR-13566590 Sao Carlos, SP - Brazil

^[2] Sao Paulo State Univ UNESP, Biodivers Dept, BR-13506900 Rio Claro, SP - Brazil

^[3] Univ Coll Cork, Sch Comp Sci & Informat Technol, Cork T12 XF62 - Ireland

• Tipo de documento: Artigo Científico
• Fonte: INFORMATION; v. 12, n. 7 JUL 2021.
• Citações Web of Science: 0

Resumo

Labeling of samples is a recurrent and time-consuming task in data analysis and machine learning and yet generally overlooked in terms of visual analytics approaches to improve the process. As the number of tailored applications of learning models increases, it is crucial that more effective approaches to labeling are developed. In this paper, we report the development of a methodology and a framework to support labeling, with an application case as background. The methodology performs visual active learning and label propagation with 2D embeddings as layouts to achieve faster and interactive labeling of samples. The framework is realized through SoundscapeX, a tool to support labeling in soundscape ecology data. We have applied the framework to a set of audio recordings collected for a Long Term Ecological Research Project in the Cantareira-Mantiqueira Corridor (LTER CCM), localized in the transition between northeastern Sao Paulo state and southern Minas Gerais state in Brazil. We employed a pre-label data set of groups of animals to test the efficacy of the approach. The results showed the best accuracy at 94.58% in the prediction of labeling for birds and insects; and 91.09% for the prediction of the sound event as frogs and insects. (AU)

• Processo FAPESP: 20/01779-5 - Biodiversidade no Antropoceno: efeito dos agroecossistemas na conservação da biodiversidade e manutenção de funções ecossistêmicas
• Beneficiário: Milton Cezar Ribeiro
• Modalidade de apoio: Auxílio à Pesquisa - Regular
• Processo FAPESP: 13/50421-2 - Novos métodos de amostragem e ferramentas estatísticas para pesquisa em biodiversidade: integrando ecologia de movimento com ecologia de população e comunidade
• Beneficiário: Milton Cezar Ribeiro
• Modalidade de apoio: Auxílio à Pesquisa - Regular