So in today's day (World Environment Day, June 5), facing the era of artificial intelligence with the rapid development of technology, let's talk about how to keep the balance between science and technology and environment.

It is a pleasure to note that more and more organizations, enterprises and individuals have joined the ranks of environmental protection. We can also feel that scientific and technological innovation is bringing more possibilities for environmental protection.

For example, in the field of identification and tracking of endangered animals, the most representative case is the deep neural network trained by Google in cooperation with the National Oceanic and Atmospheric Administration (NOAA). The model was trained with a large number of recordings collected by NOAA underwater hydrophones to identify different whale "singing" sounds and track their trajectories.

For the more complex and difficult to classify population recording processing, Google uses the existing large-scale sound to classify, converts audio data into spectrograms to train algorithm, and constructs a special humpback whale recognition device. These identification and tracking efforts can directly protect these endangered marine organisms and suggest that ships stay away from areas where they occur in large numbers to avoid damage such as entanglement or collision of fishing gear. At the same time, it can also show the changes of breeding sites, migration paths and relative abundance of these organisms and other related information to promote research.