Simplifying camera trap image analysis with AI (addaxdatascience.com)

AddaxAI has launched a groundbreaking application that leverages AI to streamline the analysis of images and videos captured by camera traps, significantly aiding ecologists in monitoring wildlife. This open-source platform allows users to perform automatic detection and identification of various species directly on their local machines without requiring an internet connection post-installation. It incorporates the MegaDetector model, which effectively filters out non-target images, thus enabling ecologists to conserve time and focus on crucial conservation efforts. Users can also verify AI predictions through a human-in-the-loop process, ensuring higher accuracy in species identification. This development is particularly significant for the AI/ML community as it exemplifies the integration of deep learning with ecological research, facilitating biodiversity monitoring at scale. The platform’s flexibility allows for processing both images and videos, with GPU acceleration for enhanced performance on NVIDIA and Apple Silicon devices. Furthermore, AddaxAI supports integration with the Timelapse Image Analyser, permitting seamless export of results, which paves the way for collaborative research and data sharing in the conservation space. This initiative, part of the OneForestVision project, aims to further enhance the open-source AI ecosystem by encouraging contributions from researchers engaged in wildlife studies.