A password will be e-mailed to you.

AI Helps Unlock Ancient Rock Art Mysteries

Day by day artificial intelligence adds newer hues of possibilities across diverse fields, unlocking new doors of opportunities for growth and progress. Now, with the latest application of artificial intelligence, the past can be studied through the eyes of the future.

AI Rock Art analysis

Archaeologists in Adelaide, Australia are harnessing artificial intelligence technology to study the development and creation of Aboriginal Rock Painting. By adding this futuristic twist to the past, archaeologists expect to unveil newer domains of historic revelations that were earlier shrouded in a vague haze of uncertainty. The AI technology used will substantially help in studying the diverse artistic styles which evolved over thousands of years.

How Does It Work

Extended field research on these paintings was deemed tough due to the challenges posed by its remote location which made access a herculean task. Added to this was the fact that most of the motifs were weathered adding another roadblock for research and analysis. It is expected that the addition of a smart technological filter will prove to be of immense help in addressing these challenges. Identification of individual artists and their styles will be made more convenient by using technologies similar to face recognition.

The archaeologists from Flinders University are focusing their work on Arnhem Land, characterized by a multitude of diverse human figure styles. The archaeologists are experimenting with smart technology to see whether it will be useful in grouping the well-defined and diverse styles.

This will also mark the first time where Australian rock art is analyzed with machine learning by utilizing a data-efficient transfer learning approach. The use of machine learning has made it possible to resolve the art styles found in Arnhem Rock Art without the pre-requisite knowledge of prior groupings. This adds to the convenience and efficiency of the research process.

The software used was trained with millions of images so that it will be made capable of identifying the differences and subtle points of distinction between aging motifs. Hundreds of different points on each rock art photo are analyzed and compared by the software. The accuracy and precision characteristic of artificial intelligence will prove to be of immense help here since the classifications of works are often difficult owing to their very vague and negligible variations. Most of these works are between 5000 -15000 years old.

The archaeologists are quite impressed by the results since, in addition to classifying them, machine learning also facilitated chronological ordering.

According to the researchers, the results derived from artificial intelligence are much more accurate since they are free from biases that arise from human pre-conceptions.

By helping to identify individual artists and styles, the technology will also help in revealing the uniqueness of rock art and its rare characteristics.,



No more articles
Send this to a friend