5G AI-IoT System for Bird Species Monitoring and Song Classification.

Identification of different species of animals has become an important issue in biology and ecology. Ornithology has made alliances with other disciplines in order to establish a set of methods that play an important role in the birds' protection and the evaluation of the environmental quality of different ecosystems. In this case, the use of machine learning and deep learning techniques has produced big progress in birdsong identification. To make an approach from AI-IoT, we have used different approaches based on image feature comparison (through CNNs trained with Imagenet weights, such as EfficientNet or MobileNet) using the feature spectrogram for the birdsong, but also the use of the deep CNN (DCNN) has shown good performance for birdsong classification for reduction of the model size. A 5G IoT-based system for raw audio gathering has been developed, and different CNNs have been tested for bird identification from audio recordings. This comparison shows that Imagenet-weighted CNN shows a relatively high performance for most species, achieving 75% accuracy. However, this network contains a large number of parameters, leading to a less energy efficient inference. We have designed two DCNNs to reduce the amount of parameters, to keep the accuracy at a certain level, and to allow their integration into a small board computer (SBC) or a microcontroller unit (MCU).

Intercomparison of Soil Moisture Retrieved from GNSS-R and from Passive L-Band Radiometry at the Valencia Anchor Station.

In this paper, the SOMOSTA (Soil Moisture Monitoring Station) experiment on the intercomparison of soil moisture monitoring from Global Navigation Satellite System Reflectometry (GNSS-R) signals and passive L-band microwave radiometer observations at the Valencia Anchor Station is introduced. The GNSS-R instrument has an up-looking antenna for receiving direct signals from satellites, and a dual-pol down-looking antenna for receiving LHCP (left-hand circular polarization) and RHCP (right-hand circular polarization) reflected signals from the soil surface. Data were collected from the three different antennas through the two channels of Oceanpal GNSS-R receiver and, in addition, calibration was performed to reduce the impact from the differing channels. Reflectivity was thus measured, and soil moisture could be retrieved. The ESA (European Space Agency)-funded ELBARA-II (ESA L Band Radiometer II) is an L-band radiometer with two channels with 11 MHz bandwidth and respective center frequencies of 1407.5 MHz and 1419.5 MHz. The ELBARAII antenna is a large dual-mode Picket horn that is 1.4 m wide, with a length of 2.7 m with -3 dB full beam width of 12° (±6° around the antenna main direction) and a gain of 23.5 dB. By comparing GNSS-R and ELBARA-II radiometer data, a high correlation was found between the LHCP reflectivity measured by GNSS-R and the horizontal/vertical reflectivity from the radiometer (with correlation coefficients ranging from 0.83 to 0.91). Neural net fitting was used for GNSS-R soil moisture inversion, and the RMSE (Root Mean Square Error) was 0.014 m3/m3. The determination coefficient between the retrieved soil moisture and in situ measurements was R2 = 0.90 for Oceanpal and R2 = 0.65 for Elbara II, and the ubRMSE (Unbiased RMSE) were 0.0128 and 0.0734 respectively. The soil moisture retrievals by both L-band remote sensing methods show good agreement with each other, and their mutual correspondence with in-situ measurements and with rainfall was also good.