Scientists have developed an AI-powered app called DinoTracker, which can identify dinosaur footprints with remarkable accuracy using artificial intelligence.
The brainchild of experts from the University of Edinburgh and Helmholtz-Zentrum in Germany, the app uses machine learning to analyze the shape of a footprint and compare it to thousands of previously identified tracks. By analyzing features such as toe spread, ground contact, and heel position, the AI system can determine how similar or different imprints are from each other.
The researchers tested their system by feeding it 2,000 unlabelled footprint silhouettes, which resulted in a remarkable accuracy rate of around 90%. The app is free to download and allows users to upload the silhouette of a footprint, explore other footprints that are similar, and manipulate the footprint to see how varying its features affects the results.
The discovery has significant implications for paleontology, particularly with regards to the origins of birds. According to Dr. Steve Brusatte, one of the co-authors of the study, the findings suggest that birds may have a much older, more ancient ancestry than previously thought, potentially dating back tens of millions of years. However, he cautioned that the evidence does not necessarily prove the existence of birds in the distant past.
One expert, Dr. Jens Lallensack from Humboldt University of Berlin, noted that while the AI system's features were not directly based on the shape of the foot itself, they still may be influenced by factors such as the type of ground or sand the dinosaur was walking through. He also suggested that the bird-like tracks may have resulted from the way a theropod sank into soft ground rather than evidence of an early bird appearance.
The DinoTracker app is set to revolutionize paleontology, enabling researchers to gain new insights into the lives and habits of ancient dinosaurs.
The brainchild of experts from the University of Edinburgh and Helmholtz-Zentrum in Germany, the app uses machine learning to analyze the shape of a footprint and compare it to thousands of previously identified tracks. By analyzing features such as toe spread, ground contact, and heel position, the AI system can determine how similar or different imprints are from each other.
The researchers tested their system by feeding it 2,000 unlabelled footprint silhouettes, which resulted in a remarkable accuracy rate of around 90%. The app is free to download and allows users to upload the silhouette of a footprint, explore other footprints that are similar, and manipulate the footprint to see how varying its features affects the results.
The discovery has significant implications for paleontology, particularly with regards to the origins of birds. According to Dr. Steve Brusatte, one of the co-authors of the study, the findings suggest that birds may have a much older, more ancient ancestry than previously thought, potentially dating back tens of millions of years. However, he cautioned that the evidence does not necessarily prove the existence of birds in the distant past.
One expert, Dr. Jens Lallensack from Humboldt University of Berlin, noted that while the AI system's features were not directly based on the shape of the foot itself, they still may be influenced by factors such as the type of ground or sand the dinosaur was walking through. He also suggested that the bird-like tracks may have resulted from the way a theropod sank into soft ground rather than evidence of an early bird appearance.
The DinoTracker app is set to revolutionize paleontology, enabling researchers to gain new insights into the lives and habits of ancient dinosaurs.