Skip to content
Home » Sandbox

Sandbox

Services

You can read about what we do at Addax Data Science below. Not much of a reader? Watch this webinar instead.

Species Recognition Models

Manually analysing camera trap images or sound files can be exhausting and time-consuming. Addax can create custom AI-driven computer vision models to automate species identification, saving you time and effort.

Early Warning Systems

Encounters between humans and wildlife are becoming increasingly common. Early warning systems use advanced cameras and AI to detect animals in real-time, reducing conflicts and enhancing safety for both humans and wildlife.

Workflow Automation

Ecological research is often repetitious. Addax can significantly streamline processes such as data extraction, validation, identification, and data flows, allowing you to focus on the insights gained and their research implications.

Statistics and Visualisation

Analysis of ecological data is challenging even for the experts. Addax can help, whether by providing a comprehensive consultation or by conducting the analysis itself—from simple comparative studies to complex scientific research.

Screenshots

Services

You can read about what we do at Addax Data Science below. Not much of a reader? Watch this webinar instead.

Statistics and Visualisation

Analysis of ecological data is challenging even for the experts. Addax can help, whether by providing a comprehensive consultation or by conducting the analysis itself—from simple comparative studies to complex scientific research.

Species Recognition Models

Manually analysing camera trap images is exhausting and time-consuming. Addax can create custom AI-driven computer vision models to automate species identification, saving you time and effort.

Workflow Automation

Ecological research is often repetitious. Addax can significantly streamline processes such as data extraction, validation, identification, and data flows, allowing you to focus on the insights gained and their research implications.

Frequently Asked Questions

Please use the following citations if you used EcoAssist in your research.

  1. van Lunteren, P. (2023). EcoAssist: A no-code platform to train and deploy custom YOLOv5 object detection models. Journal of Open Source Software8(88), 5581.
  2. Beery, S., Morris, D., & Yang, S. (2019). Efficient pipeline for camera trap image review. arXiv preprint arXiv:1907.06772.
  3. Plus the citation of the species identification model used.

EcoAssist should automatically run on NVIDIA or Apple Silicon GPU if available. The appropriate CUDAtoolkit and cuDNN software is already included in the EcoAssist installation for Windows and Linux. If you have NVIDIA GPU available but it doesn't recognise it, make sure you have a recent driver installed, then reboot. An MPS compatible version of Pytorch is included in the installation for Apple Silicon users. The progress window will display whether EcoAssist is running on CPU or GPU. Email me if you need more assistance.

It's always good practise to first run EcoAssist in debug mode, where it will print its output in a console window. That should point us in the right direction if there is an error. How to run it in debug mode depends on your operating system and can be found here. You can always email me if you if you need help with this.

Once you've opened EcoAssist in debug mode, you'll have to recreate the error so that the traceback will show up in the console window. You can copy-paste the output and email it to us, or raise an issue in the GitHub repository.

Interested in contributing to this project? There are always things to do. Do you feel comfortable handling one of the tasks listed here?

EcoAssist is an open-source project, so please feel free to fork the EcoAssist GitHub repository and submit fixes, improvements or add new features. For more information, see the contribution guidelines.

Previous code contributors can be found here. Thank you!

In previous versions of EcoAssist (v3.0 > v4.3) it was possible to train your own object detection models based on MegaDetector to detect your target species. Although this did work, it wasn't the best approach to develop a species recognition model. It required lots of training data, processing power, time, electricity and wasn't very accurate. Advancing insights revealed that better results can be obtained by using an object classification model to be used in conjunction with the results of MegaDetector. The animals will then be located by MegaDetector, and further classified by your custom model. EcoAssist > v4.2 does support the deployment of a classification model to be used in conjunction with MegaDetector, but training such a model is more complicated and hasn't been incorporated into EcoAssist > v4.4.

If you still want to use the training feature of v4.3, you can download the EcoAssist v4.3 install file below. The rest of the installation will be done as usual, as is described here.

We've placed a detailed tutorial on Medium that provides a step-by-step guide on annotating, training, evaluating, deploying, and postprocessing data with EcoAssist v4.3. You can find it here.

All EcoAssist files are located in one folder, called 'EcoAssist_files'. See these instructions on where to find it. Windows users also need to remove the 'ecoassistcondaenv' conda environments. Let me know if you need help with that.