top of page

Workshops & Courses 

研讨会与课程

W-10

Workshop 

Artificial Intelligence for Tropical Biodiversity Monitoring

TBA

Organizer(s):
Aléxia Rodrigues, Federal University of Pará
Description
Tropical ecosystems hold an extraordinary share of the planet’s biodiversity, but they are also among the most threatened. Deforestation, climate change, pollution, and invasive species are transforming forests, rivers, and coastal areas faster than scientists and local communities can monitor them. Traditional monitoring methods are often slow, expensive, and difficult to maintain over large and remote areas. At the same time, new tools based on artificial intelligence (AI), low-cost sensors, and automated cameras and recorders are becoming more accessible and powerful. These tools can help detect species, track environmental change, and provide faster information to decision-makers. This workshop will explore how AI can support environmental monitoring in tropical regions in a practical and inclusive way. We will present real examples where AI is already being used with images, audio, and other environmental data, such as camera-based insect traps, camera traps for mammals and birds, bioacoustic monitoring for frogs and birds, and image analysis for coral reefs and aquatic systems. A central example will be the IARAA project (Intelligence Artifficial for Environmental Recognition in the Amazon), which integrates automated cameras, sensors and AI-models to monitor biodiversity across remotes areas in the Amazon. We will focus on what works in the field, what still does not work well, and what is needed to make these tools more useful and equitable. The goals of the session are to: (1) introduce participants to the basic ideas behind AI-assisted monitoring without assuming advanced technical knowledge; (2) show concrete workflows from data collection to simple outputs that can support management and conservation; (3) discuss the social, ethical, and logistical challenges of using AI in tropical regions, including questions of data ownership, participation of Indigenous and local communities, and unequal access to technology; and (4) build connections among researchers, students, practitioners, and community representatives who are interested in collaborating. By the end of the workshop, participants will have a clearer understanding of how AI tools can be integrated into ongoing or future monitoring efforts in tropical ecosystems, what resources (data, training, partnerships) are needed, and how they can join or initiate collaborative projects that use AI to support conservation and environmental justice in the tropics.
Program Outline

1. Welcome and overview (10 min) Brief introduction of organizer and participants. - Presentation of the workshops goals: understanding basic AI concepts and experiencing hands-on model training. - Short outline of the session structure (theory + practice + discussion). 2. Introdutory talk: AI concepts for biodiversity monitoring (20min) - Simple explanation of key concepts and their parameters: classification, object detection and segmentation, using visual exemples from ecology and conservation. - Overview of typical dara sources (images, videos, sensor data) and why AI is useful in tropical contexts (data gaps, remote areas, limited staff) 3. Hands-on session I - Classification and detection in Google Colab (30min) - Guided, step-by-step use of Google Colab to train models using prep-prepared datasets. - Participants follow along with theur own laptops. 4. Hands-on session II - Segmentation (20min) - Demonstration of image workflows in Google Colab (e.g., separating organisms from background, mapping structures). - Discusion of how different model types answer different ecological questions. 5. What works and whats does not: Lessons from the field and from IARAA (25min). - Discussion of pratical limitations: data quality, annotation effort, computing resources and biases. - Presentation of the IARAA project as a concrete exmple of how AI is used to monitor biodiversity in tropical landscapes, incluinding remote sensing, automated cameras and integration with local contexts. - Highlighting sucesses, failures and ongoing challenges.

Materials that participants need to bring:

A laptop with a modern web browser (Chrome or Firefox is recomended). 

bottom of page