Scientists Use AI to Measure Hummingbird Wingbeats

Jul 30

Hummingbirds are famous for their incredibly fast wingbeats, often exceeding 40 to 80 beats per second depending on the species. These rapid movements allow hummingbirds to hover in place, fly backward, and maneuver with extraordinary precision.

Studying hummingbird wing motion, however, has always been a challenge. Because their wings move so quickly, traditional observation methods often struggle to capture the full complexity of their flight.

A recent scientific study introduces a new method for analyzing hummingbird flight using computer vision and machine learning. By applying artificial intelligence to video recordings, researchers were able to estimate hummingbird wingbeat frequency without manually labeling the footage. This new approach may help scientists study hummingbird flight more efficiently and at larger scales than previously possible.

The Challenge of Measuring Wingbeats

Hummingbird wings move so rapidly that even high-speed cameras can produce massive amounts of data that must be analyzed frame by frame.

Traditionally, researchers estimate wingbeat frequency by:

recording high-speed video
manually tracking wing movement
measuring timing between wing strokes

While effective, this process is extremely time-consuming and requires extensive manual work.

Because of these challenges, scientists have been searching for automated ways to analyze hummingbird flight.

Using Computer Vision to Study Flight

In this study, researchers developed a computer vision approach capable of estimating hummingbird wingbeat frequency directly from video footage.

The system analyzes visual patterns in the video and detects rhythmic motion associated with wingbeats. Instead of requiring researchers to manually label each frame, the algorithm learns to recognize the movement patterns on its own.

This approach is known as no-labeling learning, meaning the system can analyze data without needing large manually annotated datasets.

Why Artificial Intelligence Helps

Machine learning tools are becoming increasingly valuable in biological research because they can process large datasets quickly and consistently.

Using computer vision to analyze hummingbird flight offers several advantages:

faster analysis of large video datasets
reduced need for manual labeling
improved consistency in measurements

These tools allow researchers to study flight behavior more efficiently and could open the door to new large-scale studies of hummingbird movement.

Expanding the Tools of Hummingbird Research

The use of artificial intelligence in this study demonstrates how modern technology is transforming wildlife research. By combining high-speed video with machine learning, scientists can analyze complex biological motion in ways that were previously difficult or impractical.

For hummingbirds, this approach may help researchers better understand flight mechanics, energy use, and behavioral differences among species.

As new analytical tools continue to develop, studies like this show how technology can deepen our understanding of one of nature’s most remarkable flyers.

Research Credit

This article summarizes findings from the following scientific study:

Maria Ximena Bastidas-Rodriguez, Ana Melisa Fernandes, María José Espejo-Uribe, Diana Abaunza, Juan Sebastián Roncancio, Eduardo Aquiles Gutierrez-Zamora, Cristian Flórez Pai, Ashley Smiley, Kristiina Hurme, Christopher J Clark, Alejandro Rico-Guevara (2025).
Estimating Wingbeat Frequency of Hummingbirds Using a No-Labeling Learning Computer Vision Approach.
Integrative and Comparative Biology, Volume 65, Issue 1, July 2025, Pages 127–138,

Read the Original Paper

Readers interested in the full methodology and technical details can consult the original research paper.

Frequently Asked Questions

Research on hummingbird flight often raises questions about how scientists measure wing motion and why these measurements are important. Here are some common questions related to hummingbird wingbeats and flight analysis.

Depending on the species, hummingbird wingbeats typically range from 40 to more than 80 beats per second.
Because hummingbird wings move extremely fast, researchers must use high-speed video to capture the motion accurately.
Computer vision is a type of artificial intelligence that allows computers to analyze visual information such as images or video.
AI can analyze large amounts of video data quickly and detect patterns in wing motion that would otherwise require manual analysis.
Wingbeat analysis helps scientists understand hummingbird flight mechanics, energy use, and aerodynamic performance.