WiFi everywhere with Google Loon
Helium balloons are carrying WiFi connectivity to developing regions lacking network infrastructure
Researchers are developing new software to translate the language of prairie dogs, marmosets, and dolphins, among other species.
The applications of artificial intelligence are covering an increasing number of fields. And linguistics would be one of the most promising areas of research. For instance, there are already plenty of neural network projects aiming to improve the quality of automated translation. Recently, we covered the analysis of the Indias archive through artificial intelligence. This Artificial Intelligent project established that Spain was probably the discoverer of Australia, besides finding several unheard-of shipwrecks. In fact, if you have reached this article through a search engine, it is perhaps through an AI algorithm. Artificial intelligence is remarkably good at finding patterns and establishing semantic connections. And communication with animals could well be its next breakthrough.
Currently, several research teams are working on the development of new artificial intelligence tools to enable a better understanding of animals from different species. Professor Slobodchikoff realized that prairied dogs were able to describe the size and color of their predators in a sophisticated way. I.e., they could even distinguish humans by the color of their clothes. Discovering those linguistic skills drove Slobodchikoff to found Zoolinguam a startup focused on developing translation software based on AI for pets like dogs.
Scientists from MIT have carried out another experiment in the field of Artificial Intelligence, this time developing software to interpret the chirps and sounds emitted by marmosets. This type of monkeys uses between ten and fifteen different calls and warnings. The program can isolate each expression from the background noise. Thanks to the use of artificial intelligence, the program can assess the meaning of each sound with a 90 percent accuracy.
Besides terrestrial mammals, scientists have studied the communication of cetaceans for years. Gavagai AB, a Swedish startup specialized in software to analyze human languages, has partnered with the KTH Royal Institute of Technology. The goal is to employ artificial intelligence to understand the language of dolphins. They hope to have a functional version by 2021.
Nevertheless, communication, both for human beings and animals, is not limited to sounds and phonemes.
Another way of understating animals is their body language and facial expressions. And, while artificial intelligence has proved its efficiency with linguistics, image recognition is another of its key strengths. Scientists from the University of Cambridge have developed new software that can analyze the face of sheep and detect signs of pain, which so far only a trained vet had been able to do. To achieve their goal, they analyzed around five hundred images, looking for variations in the narrowing of the eyes, nostril-deformation, or the rotation of each ear. The angle of the ears was one of the most telling cues. The AI software achieved a 67 percent precision, similar to human visual analysis, although it could improve with further training. The researchers are confident that the results will soon apply to other species.