@article{fdi:010092133,
  title = {{I}mproving {CNN} fish detection and classification with tracking},
  author = {{Z}ouin, {B}. and {Z}ahir, {J}. and {B}aletaud, {F}. and {V}igliola, {L}aurent and {V}illon, {S}{\'e}bastien},
  editor = {},
  language = {{ENG}},
  abstract = {{T}he regular and consistent monitoring of marine ecosystems and fish communities is becoming more and more crucial due to increasing human pressures. {T}o this end, underwater camera technology has become a major tool to collect an important amount of marine data. {A}s the size of the data collected outgrew the ability to process it, new means of automatic processing have been explored. {C}onvolutional neural networks ({CNN}s) have been the most popular method for automatic underwater video analysis for the last few years. {H}owever, such algorithms are rather image-based and do not exploit the potential of video data. {I}n this paper, we propose a method of coupling video tracking and {CNN} image analysis to perform a robust and accurate fish classification on deep sea videos and improve automatic classification accuracy. {O}ur method fused {CNN}s and tracking methods, allowing us to detect 12% more individuals compared to {CNN} alone.},
  keywords = {marine ecosystems ; convolutional neural networks ({CNN}s) ; {BRUVS} video ; data ; fish classification ; automatic processing ; tracking},
  booktitle = {},
  journal = {{A}pplied {S}ciences},
  volume = {14},
  numero = {22},
  pages = {10122 [13 p.]},
  year = {2024},
  DOI = {10.3390/app142210122},
  URL = {https://www.documentation.ird.fr/hor/fdi:010092133},
}