Audio-based insect monitoring

1 . A method, comprising:

receiving, by an insect activity prediction system, audio data associated with at least one microphone in contact with a surface of a plant;

receiving environmental data associated with a location of the plant at a time of collecting the audio data;

determining a validity status of the audio data based on the environmental data;

validating the audio data based on determining that the validity status confirms one or more environmental conditions, wherein the one or more environmental conditions comprise one or more of a wind level below a wind threshold and a precipitation activity below a precipitation threshold;

inputting, by the insect activity prediction system, the audio data into an insect activity model to generate an insect activity prediction;

generating, by the insect activity prediction system using the insect activity model, an insect activity prediction of a presence of insects and at least one species of the insects based at least in part on the audio data and the validity status of the audio data; and

outputting the insect activity prediction to a user interface.

2 . The method of claim 1 , wherein the audio data is captured by a data collection device comprising a processing system through one or more audio interfaces configured to capture audio input from the at least one microphone.

3 . The method of claim 2 , wherein two or more of the at least one microphone capture a plurality of audio streams associated with the plant.

4 . The method of claim 2 , wherein two or more of the at least one microphone capture a plurality of audio streams associated with at least two different plants.

5 . The method of claim 1 , wherein the environmental data is captured by a data collection device that captures the audio data and/or received from a data source that differs from the data collection device that captures the audio data.

6 . The method of claim 1 , wherein the insect activity model comprises one or more machine learning models trained to distinguish between audio characteristics of insect chewing, insect sucking, insect tapping, and insect movement on or within plant leaves, flowers, fruit, stems, branches, stalks, petioles, buds, and/or roots at one or more insect life stages.

7 . The method of claim 1 , wherein the insect activity model comprises a first convolutional neural network trained to identify the presence of the insects and a second convolutional neural network trained to identify at least one species of the insects.

8 . The method of claim 1 , wherein the insect activity model is selected from a plurality of models trained for various plant and insect types, and the insect activity model is selected based on a geographic location of the plant.

9 . The method of claim 1 , further comprising:

combining the insect activity prediction for the plant with one or more insect activity predictions for a plurality of plants to produce an insect map for one or more geographic regions.

10 . A system comprising:

a memory system having a plurality of computer readable instructions; and

one or more processors configured to execute the computer readable instructions to perform a plurality of operations comprising:

receiving audio data associated with at least one microphone in contact with a surface of a plant;

receiving environmental data associated with a location of the plant at a time of collecting the audio data;

determining a validity status of the audio data based on the environmental data;

validating the audio data based on determining that the validity status confirms one or more environmental conditions, wherein the one or more environmental conditions comprise one or more of a wind level below a wind threshold and a precipitation activity below a precipitation threshold;

inputting the audio data into an insect activity model to generate an insect activity prediction;

generating, by the insect activity model, an insect activity prediction of a presence of insects and at least one species of the insects based at least in part on the audio data and the validity status of the audio data; and

outputting the insect activity prediction to a user interface.

11 . The system of claim 10 , wherein the audio data is captured by a data collection device through one or more audio interfaces configured to capture audio input from the at least one microphone.

12 . The system of claim 10 , wherein the insect activity model comprises a first convolutional neural network trained to identify the presence of the insects and a second convolutional neural network trained to identify at least one species of the insects.

13 . The system of claim 10 , wherein the insect activity model is selected from a plurality of models trained for various plant and insect types, and the insect activity model is selected based on a geographic location of the plant.

14 . The system of claim 12 , wherein the computer readable instructions are executable by the one or more processors to perform the operations further comprising:

combining the insect activity prediction for the plant with one or more insect activity predictions for a plurality of plants to produce an insect map for one or more geographic regions.

15 . A computer program product comprising a non-transitory computer readable storage medium embodied with computer program instructions that when executed by one or more processors cause the one or more processors to perform a plurality of operations comprising:

receiving audio data associated with at least one microphone in contact with a surface of a plant;

receiving environmental data associated with a location of the plant at a time of collecting the audio data;

determining a validity status of the audio data based on the environmental data;

validating the audio data based on determining that the validity status confirms one or more environmental conditions, wherein the one or more environmental conditions comprise one or more of a wind level below a wind threshold and a precipitation activity below a precipitation threshold;

inputting the audio data into an insect activity model to generate an insect activity prediction;

generating, by the insect activity model, an insect activity prediction of a presence of insects and at least one species of the insects based at least in part on the audio data and the validity status of the audio data; and

outputting the insect activity prediction to a user interface.

16 . The computer program product of claim 15 , wherein the insect activity model comprises a first convolutional neural network trained to identify the presence of the insects and a second convolutional neural network trained to identify at least one species of the insects.

17 . The computer program product of claim 15 , wherein the environmental data is captured by a data collection device that captures the audio data and/or received from a data source that differs from the data collection device that captures the audio data.

18 . The computer program product of claim 15 , wherein the insect activity model comprises one or more machine learning models trained to distinguish between audio characteristics of insect chewing, insect sucking, insect tapping, and insect movement on or within plant leaves, flowers, fruit, stems, branches, stalks, petioles, buds, and/or roots at one or more insect life stages.

19 . The computer program product of claim 15 , wherein the insect activity model is selected from a plurality of models trained for various plant and insect types, and the insect activity model is selected based on a geographic location of the plant.

20 . The computer program product of claim 15 , wherein the computer program instructions are executable by the one or more processors to perform the operations further comprising:

combining the insect activity prediction for the plant with one or more insect activity predictions for a plurality of plants to produce an insect map for one or more geographic regions.