English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Underwater acoustic cameras are high potential devices for many applicationsin ecology, notably for fisheries management and monitoring. However how toextract such data into high value information without a time-consuming entiredataset reading by an operator is still a challenge. Moreover the analysis ofacoustic imaging, due to its low signal-to-noise ratio, is a perfect trainingground for experimenting with new approaches, especially concerning DeepLearning techniques. We present hereby a novel approach that takes advantage ofboth CNN (Convolutional Neural Network) and classical CV (Computer Vision)techniques, able to detect a generic class ''fish'' in acoustic video streams.The pipeline pre-treats the acoustic images to extract 2 features, in order tolocalise the signals and improve the detection performances. To ensure theperformances from an ecological point of view, we propose also a two-stepvalidation, one to validate the results of the trainings and one to test themethod on a real-world scenario. The YOLOv3-based model was trained with dataof fish from multiple species recorded by the two common acoustic cameras,DIDSON and ARIS, including species of high ecological interest, as Atlanticsalmon or European eels. The model we developed provides satisfying resultsdetecting almost 80% of fish and minimizing the false positive rate, howeverthe model is much less efficient for eel detections on ARIS videos. The firstCNN pipeline for fish monitoring exploiting video data from two models ofacoustic cameras satisfies most of the required features. Many challenges arestill present, such as the automation of fish species identification through amulticlass model. 1 However the results point a new solution for dealing withcomplex data, such as sonar data, which can also be reapplied in other caseswhere the signal-to-noise ratio is a challenge.

A deep neural network for multi-species fish detection using multiple acoustic cameras