2019 Hackweek Projects
Oceanhackweek is not just about learning, it is also about using what is learned to create exciting new tools in collaboration with others. We strive to foster an inclusive and supportive learning environment based on the shared interests of all participants, and are proud to share outcome of the hack projects here.
YouTube links to the final presentations are coming soon.
Project: TMIFVP
Template Matching to Identify Features in Vertical Profiles, or TMIFVP, is a tool designed to match and quantify the difference between a template profile and user-submitted data, clearly highlighting features for quality matching scores and location of problematic or interesting attributes. More broadly, this pipeline could be used to analyze and compare many types of graphs and profiles, including but not limited to: Michaelis-Menten population curves, Metabolomic and proteomic spectrographic peaks, and environmental time series data to identify outliers.
Github Page: https://github.com/oceanhackweek/ohw19-projects-TMIVP
Presentation: video
Project: Trackpy
This project allows Python beginners a way to easily link and visualize animal movement data with sensed oceanographic data and databases of mesoscale oceanographic features.
Github Page: https://github.com/oceanhackweek/ohw19-projects-Trackpy
Presentation: video
Project: Isopy
Isopy is a python package that allows users to calculate isosurfaces, and variables along isopycnals, from 3d datasets.
Github Page: https://github.com/oceanhackweek/DataAccess/tree/master/isopy
Presentation: video
Project: Underwater Currency Counter
Called, “a Semi Supervised Annotation Tool: Kickass AI to count sand dollars”, the Underwater Currency Counter is an effort to train a deep object recognition algorithm to count sand dollars from benthic habitat imagery data collected from ROVs.
Github Page: https://github.com/oceanhackweek/ohw19-project-computer_vision_club
Presentation: video
Project: Co-Locators
This team aims to provide a way to Co-locate oceanographic data by establishing constraints. The tools they developed allow users to provide geospacial bounds in a temporal range to get an aggregated response of all available data.
Github Page: https://github.com/oceanhackweek/ohw19-project-co_locators
Presentation: video
Project: 21st Century Prediction of Fish Larvae Catch Using ML
This project is a machine learning effort to predict fish larvae catch numbers.
Github Page: https://github.com/oceanhackweek/ohw19-project-CMIP6-larvae-ML
Presentation: video
Project: Modeling Volcano Deformation at Axial Seamount
A Jupyter notebook intended to teach and demonstrate to undergraduates about volcano and seafloor deformation, and how previous volcanic eruptions of Axial Seamount were used to forecast future eruptions. This notebook also allows students to forecast the next eruption themselves!
Presentation: video
Project: DataStreamSync
The objective is to create an interface that allows to visualize the spectrogram and waveform of the fish sounds detected as well as the video and sonar data at the time of the detection.
Github Page: https://github.com/oceanhackweek/ohw19-projects-DataStreamSync
Presentation: video
Project: MLQC
Applying machine learning techniques to QC timeseries data, and compare against established algorithmic approaches (e.g. QARTOD).
Github Page: https://github.com/oceanhackweek/ohw19-project-mlqc-for-timeseries
Presentation: video
Project: Working with Chlorophyll Data from the Cloud
Notebooks demonstrating how to work with chlorophyll data including how to remotely access chlorophyll data from OceanData.sci, how to read the data, and how to compute local gradients in the measured values.
Github Page: https://github.com/oceanhackweek/DataAccess/tree/master/Chlorophyll
Presentation: video
Project: Amazon Fires
A learning tool used to show Amazon data including bushfires, daily surface temperature over 10 years, El Nino events, and signals in river discharge.
Presentation: video