Katja Ovchinnikova

9 days ago by Andrei Rusu & Katja Ovchinnikova — 11 min read

An End-to-End ML at Edge Pipeline for Target Detection in Hydroacoustic Data

A detailed outline of how we implemented an Edge‑to‑Cloud ML-based pipeline where we trained and deployed a model for semantic segmentation of echograms.

Training a Deep Learning Model for Echogram Semantic Segmentation

Andrei Rusu & Katja Ovchinnikova — 10 min read

In this tutorial we build a deep‑learning pipeline for echogram segmentation using open‑source tools. Echograms are two‑dimensional plots of acoustic echo intensity versus time and depth recorded using sonar instruments, in our case echosounders.

Using Copernicus Satellite Observations to Improve Ocean Acoustic Measurements

Andrei Rusu & Katja Ovchinnikova — 8 min read

Sound travels faster or slower depending on temperature, salinity and pressure, so an accurate sound speed profile (SSP) and absorption coefficient of seawater are essential for converting raw echosounder signals into meaningful depths and calibrated backscattering strengths.

Python Jupyter Physical Oceanography SSP Copernicus satellite observations

An End-to-End ML at Edge Pipeline for Target Detection in Hydroacoustic Data

Training a Deep Learning Model for Echogram Semantic Segmentation

Using Copernicus Satellite Observations to Improve Ocean Acoustic Measurements

Stay in the loop